tv Facebook Google and Twitter CE Os Testify on Combating Online... CSPAN March 25, 2021 3:18pm-5:33pm EDT
a business decision. different businesses and services will have different policies, some more liberal than others, and we believe it's critical this variety continues to exist forcing every business to perform the same diminishes free market ideals. if instead we woke up tomorrow and decided to ask the government what content to take down or leave up, we would not be left with a service to question the government, and it's against the right of individuals. this would have enormous requirements on businesses that would further intrench only those that could afford to use it. so how do we resolve these two viewpoints? one way is to create shared protocols. social media has proven itself important enough to be worthy of
racial and ethnic minority groups, americans are facing growing online environments. in october 2020, facebook temporarily has civic groups on the platforms. a change now made and we stormed the capitol on january 6th. i seriously question facebook's commitment to actually stopping extremism. in a recent investigative report, a former facebook a.i. researcher said he and his team conducted study after study, confirming the same basic idea.
models that maximize engagement, increase polarization. and the more engagement the generally receives. engagement is the key to facebook's growth and success. and the stock markets rewarded you for it. even as you've been criticized for promoting extremism and racist content, including in a 2020 facebook civil rights audit. the two seem to go hand in hand as facebook was also the most sited social media site in charging documents that justice department filed against the capital insurrectionshes. mr. zuckerberg, do you still maintain the more likely user content is to violate standards, the more engagement it will receive? yes or no? >> thanks for raising this because i think there's been a bunch of inaccurate things about this shared today. there seems to be a belief -- sorry.
this is a nuance topic. so, if you're okay with it, i'd like to -- >> make it sure but -- >> people don't want to see misinformation or divisive content on our services. people don't want to see click bait and things like that. while it may be true people may be more likely to click on it in the short term, it's not good for the business or product or community for the content to be there. it's not what people want and we run for the long term, with a view of 10 or 20 years from now and i think we're highly aligned with our community in trying to not show people the content that's not going to be meaningful for them. >> i only have two minutes left. do you agree with the statement in the most recent 10k filing
that it's our ability to add and retain users and maintain levels of user engagement with our products? yes or no, please. >> i think that's generally right. for any product, the unlt to build something people like and use is something that is a risk. we can't do that. >> so, do you still agree with the statement of your cfo, on the recent earnings call, wouldn't effect your engagement? yes or no? >> congresswoman, there are so many different parts of the service, that i think it's probably right not recommending political or civic groups probably isn't going to meaningfully decrease engagement. we've taken abilitier steps, including reducing viral videos by 50 million hours of watching a day, which have had a meaningful impact on engagement. we do that because it helps make the service better and make people like it more, which i think will be better for our
community and the business as a long term. >> i'm sorry to do this in five minutes. but give me your promises in the fall, the events on january 6th and true incentives, as you admit, i find it difficult to take your assurances seriously. i believe they should have access to facebook and other large media platforms algorithms, not just for groups but any relevant feature that could be exploited or exploit data from the company to support extremism and i support legislation to do so. mr. zuckerberg, given your inability to manage your algorithms or unwillingness to reduce controversial content, are you opposed to a call allowing regulators that results in the promotion of harmful disinformation and extremists content? >> congresswoman, while i don't
necessarily agree with the characterization is an important thing. we have people working on figuring out how to do this. one of the nuances in complexity is it's hard to separate out the algorithms verses people's data, which goes into that too, make decisions and the data's private. so, it's difficult to make that public and transparent. i think this is an important area of study in learning how to audit and make transparent. >> the chair recognizes mr. duncan for five minutes. >> let me just say democrats repeating disinformation is in a hearing of misinformation is irony at the most. my thoughts and prayers go out to the families that were impacted, it is not a hate crime and to say so is misinformation.
is it okay for a white male to tweet a kkk klansman hood to a black woman? >> no. >> candice was sent a tweet from a white liberal depicting a kkk hood and your support center that racist harassment of a -- harassment of a conservative didn't violate your terms of service. what do you say about that? >> we removed that tweet. >> thank you for doing that. >> and syrian refugee, a biden-supporting muslim, allegedly murdered 10 people colorado. they said referring to this gentleman as a white christian terrorist wasn't a violation of your misinformation policy? >> i don't know that case but we can follow up with you on it. >> thank you.
and your promises from the last hearing that you'll work on this and make it better ring hollow, so i'll ask that you do. you've taken down accounts of conservative and pro-life groups. at the same time liberals and terrorists continued to have unfettered access on twitter. you're able to take down an account of a sitting united states president, while he was still president, but allow state sponsors of terror to use twitter. including assad of syria. you act like judge and jury but in the hide behind the liability protections of communication decency act, which congress set up to foster a free and open internet. you think you're above the law because, in a sense, congress gavl you that power but congress gavl you that to one end. that was the protection of innocent children. catherine knocked it out of the park today, hammering the point where children are vulnerable.
let's look at john doe verses twitter case on going right now. a teenage boy, a victim, had images go viral and became bullied to point of being suicidal. you failed to take them down. his mother contacted you to alert you and again you failed to take them down. they called the police and followed up with you in a police report. your support center told the family that, after review, the illegal video was not a violation of your terms of service. it proved over 167,000 views. it took a threat to get twitter take down the video. even then, you continue to share sexually explicit videos of minors who fear violation of the law and in fear of violation, use section 230 of the
communications decency act, as they were passed. and it's better to be a ped file, pornographer or a state sponsor of terror, than it is to be a conservative, even a conservative president. you abused the shield we gave you to protect children and used it to silence conservatives instead. your privileges are far too numerous to share and to few to be removed. those who fail to score points, than protect children, my colleagues have been asking if you continue to receive immunity under section 230. no, you don't. you think you do but you don't because you continue to do a disservice to that law and its intent. united states constitution has a
first amendment, protecting the speech of users of your platform instead of treating them like hostages and forcing things through algorithms, to lead them down a path. american people are tired of you abusing your rights, abandoning their values. one of the christian leaders you banned, mr. dorsey. had as her last post a scripture verse you took down. seek peace and pursue it, rather than silence that wise advice, i strongly suggest you follow it. i've heard a lot of stuff on this hearing today about 230 protections. i challenge my colleagues to really get serious about doing something about this liability shield so we have a fair and free internet and people aren't censured. with that, i yield back. >> the chair recognizes miss kelly for five minutes.
>> thank you, mr. chair, thank you to the witnesses for testifying today. the business model for your platforms is quite simple. keep users engage. the more time they spend on social media, the more data harvests and targeted, ads sold. social media platforms amplify contend that gets attention. cat videos or vacation pictures but too often it means content that's incendiary, contains conspiracy theories or violence. algorithms can actively funnel users from the mainstream to the fringe, subjecting users to more extreme content, all to maintain user engagement. this is a fundamental flaw in your business model, that mere warning labels and even content moderation cannot address. and your company's insatiable desire to maintain user engagement will continue to give such content a safe haven to
improve the bottom line. i'd like to ask my first question all of the witnesses. and on platform? just say yes or no. starting with mr. dorsey, yes or no? >> no, it's not our business. >> mr. zuckerberg? >> no, congresswoman, i don't think we profit from it. i think it hurts our service. >> congresswoman, certainly not our intent. >> since you all said no, can you please provide how you manage to avoid collecting revenue from ads served on such content? so, i'll be expecting that. there's a difference between the conversation in the living room
and discourt and jurying hate crimes and violent. do you acknowledge that such content is having a harmful effect on minority and communities of color? yes or no. i don't have a lot of time. so, yes or no? dorsey. >> yes, i think that's right. if your financial incentive could lead to the creation of a system that promotes commercially-charged content that's been harmful, do you believe you will always need to play more on different topics?
mr. zuckerberg? >> i think we can take systemic actions that help to reduce a large amount but there will always be some content that gets through the systems that we will have to react to. >> mr. dorsey? >> i agree with mark. our models have missing things. >> i concur with what mark and jack said. there are many and we are very vigilant. >> so, more transparency and research into the a.i. models you use is needed. i understand they are constantly
volving and prupriatory, however, those obstacles must not be insurmountable. would you agree to some type of test to evaluate your procedures and technology that despaired impact? and would you set minimal standards set by the government? you have 44 seconds. >> you're not calling us. but yes. we're interested in opening all this up and going a step further and having protocol. i don't think that should be government driven but should be open and transparent. >> i agree this is an area where research would be helpful and i think some standards, especially among the civil rights community would be helpful guidance for the companies . >> congresswoman, we work with
collaboration be had. >> all right. >> thank you very much. many of the questions deal with personal arms. but they're a long term economic and security content. we host this large presence, including civilian support government. one of these is applied research associates, which is doing great work in the work of artificial intelligence and machine learning. i agree with our nation's top national security experts on the critical importance of the united states maintaining its competitive edge in a.i. and share the concern of erick schmidt, who warned of the grave
consequences if we move to china. and lay out clear a.i. strategy. we all recognize that china is not a good place to do business. evidenced by the fact that all of you are respective, main products and services are banned there. it's clear that the influence on the chinese commerce party includes the entire corporate structure. integrated with party's leadership into all aspects of corporate government. let's be clear with each other. it's impossible to do business with china without either directly or indirectly aiding the chinese commerce party. and each of your business models involve collecting data from individuals who use your product. and using that data for some other purpose. i'm deeply concerned with google's pursuit of and investment in artificial
intelligence research in china. first and foremost, could you assure americans that their personal data, regardless of how you think you have identified, data you collect in, which is central to their algorithms, is not used in your artificial intelligence collaboration? >> congressman, i want to correct any misperceptions. we do not have an a.i. center in china now. we have a limited work on own source projects, primarily open source projects on k through 12 education. a handful of employees. we don't have that anymore. compared to previous years, we don't offer core services in china. >> i'm going to have to refrain my time, because it's limited. i want your team to follow up with you because i'm somewhat skeptical and i think you had three centers there in china and
i want to know more about what they're doing and what material they're using. i'm not suggesting that doing business in a country means that you endorse all their policies. as a former businessman myself, i know politics. however, google's own list says it will not collaborate on technologies that gather and use surveillance, to widely accepted principals of international and human rights. we know the chinese communist party is using artificial intelligence to suppress the pro-democracy movement in hong kong, as well as using it in genocidal crimes, including organ harvest. once again, can you be sure that none of the work you're doing in collaboration with chinese government is not aiding them?
>> congressman, happy to follow up. the limited work is primarily around open source projects and very happy to engage and specifically follow up on what we do. >> i think that's great and i know i'm running out of time but i ask that we continue this dialogue and i think google would be very well served by promoting grader transparency in all of its actions regarding artificial intelligence in china. your customers have a right to know about this. 2018, former ceo of google cloud noted we believe the uses of our cloud and artificial intelligence include the overwhelmingly positive for the world and we recognize we cannot control our downstream purposes of the technology. a good place to start is to end the research relationship with
china. so, with that, thank you all to members of the witness panel and mr. chairman, i yield back. >> gentleman yields back. chair recognizes mr. mckeechen for five minutes. >> thank you, mr. chairman, and to you. and thank you for convening today's hearing and for our witnesses joining us. in july of last year, i heard 30 of my colleagues asking what you would doing to help disinformation on your platforms. as my colleagues fully expressed in the letter, climate change is a real threat and the spread of disinformation on the platform is undermining that. the world health organization estimates a number will increase over the years.
and why do your platforms have climate change disinformation? and mr. zuckerberg, facebook recently implemented the climate change center, with climate change facts from researchers and organizations. are you able to smear data on how disinformation is and how much climate change information center has reduced it? >> sure, thanks, congressman. our approach to disinformation is climate change is a big issue. so, i agree with your point here. we take a multi-pronged approach. one is to show authoritative information, which is what the climate information center does. and we try to reduce the spread of disinformation through this broad fact-checking program that
we have, in which, one of the fact checkers -- there is a specifically focussed on science, feedback and client feedback type of issues. overall, i'd be happy to follow up and share more details. but this is an area that i agree is extremely important and needs multiple tactics to address. >> thank you and it's my understanding it's modelled after your 2019 information center. however, different standards do apply. and climate change verses covid-19. the same standards that it does on covid-19. >> and part of the information center was based on our work on the information center and election center. in terms of how we treat this
information overall and we divide the misinformation into things that could cause imminent harm, in which covid misinformation that may lead someone to get hurt, falls into the category of imminent physical harm. w we take down that content. things that are false but may not be imminent harm are labelled and leave them up. that's the broad approach that we have. and that sort of explains the differences between some of the different issues and how we approach them. >> i hope i'm pronouncing that correctly. protection tools linking viewers to sources at the climate center. you're restricted but not released people, such as university and how they reduce
information on climate change. are you not concerned that, by restricting the videos, that people who are determined to find those videos, will indeed find them and share them with ugers? >> congressman, it's an incredibly important area. we rely on authoritative information, both by sharing information panels, as well as raising scientific content and journalistic content. so, our algorithms rank those higher. and you know, obviously it's an area where there is a range of opinions people can expect. we have clear policies, and if violated we remove. if it is not deemed to be high quality, we don't recommend the content and that's how we approach it. and as a company, we lead in
sustainability. we operate on a carbon-free basis by 2030. it's an area we're in significantly. >> well, thank you. i've run out of time. apologize to you. mr. chairman -- >> gentleman yields back. and recognizes mr. curtis for five minutes. >> thank you, mr. chairman, and thank you to our witnesses. my first comment is to point out that in her 2019 presidential campaign, senator elizabeth warren democrat, called for the breaking up of your companies. several weeks ago, in a speech at cpac, said the big tech companies should be broken up. i don't want to point out the irony of him quoting elizabeth warren at cpac.
unfortunately, the few tools we have in our bag are breaking up and sanctions. you state you cannot control and do not take responsibility for content. the community standards document, which is sited as why content is says you sometimes make moderation decisions based on what's best for public discourse. i know in your testimony, you said companies need to earn their liability protections. that doesn't address the concerns they have on past or current views on what is or is not acceptable. how do you plan to not take responsibility and maintain liability protections on your site but at the same time state your moderator content, based on the best interest that appears to be two sided. >> congressman, thanks.
people use our services to share and send messages billions of times a day. it would be impossible for us to scan or understand everything that was going on and i don't think that our society would want us to take the steps that would be necessary to monitor every single thing. i think we would think that would infringe on our freedoms. broadly, i think it's impossible to ask people to take responsibility for every piece of content that we post. at the same time, i do think we should expect large platforms to have effective systems for being able to handle, broadly speaking, categories of content that are clearly illegal. we talk said about child exploitation and opioids and sex trafficking and things like that. and i think it's reasonable to expect that companies have systems that are broadly
effective, even if they're not perfect and there are some pieces of content that get through. just like no police department -- >> i would love more time on this with you. let me ask you. utt is known for our start-up community. some view regulation with skepticism because larger companies tend to deal with regulation much better than small companies. if you think to your early start-up phase at facebook, what challenge do you see for start ups to compete and what should congress consider, as we look at regulation, that could be a barrier for companies that just might be your future competition? >> i think this is a really important point, whenever we're talking about regulation. i want to be clear that the recommendations i'm making for section 230, i would only have apply to larger platforms. i think it's really critical that small platform, the next
student in a dorm room or garage needs to have a relatively low -- as low as possible, regulatory birdn to get to the scale where they can put those systems in place. i think that's a really important point to make. i think that goes for the content discussions around 230 and probably applies to the privacy law but i hope congress will pass and this year or next year, to create federal, u.s. privacy standards. and i also think that we should be exploring, proactively requiring things like data portability that would make it easier for people to take data from one service to another. >> thank you. i got just a few seconds left. and this is a little off topic, so, i'm going to ask this question and submit it for the record. almost a decade ago your company started google fire bird and
introduced google speed and sadly, it seems like your efforts to do this were slowed down or stopped by excessive government regulations. i'd like you to share, off the record, and i'll submit it for the record, why you think it's so hard to expand internet across the country. thank you and i yield my time. >> gentleman yields back. the chair recognizes mr. soto for five minutes. >> thank you, mr. chairman. when television, radio, traditional newspapers and privacy to spread laws, they can be sued and held liable. but pursuant to 230, you can't. but at 8/19/96 -- and americans are getting hurt or killed. and the reason is your algorithms.
i want you to know i was held captive during the capitol insurrection. i was surrounded by domestic terrorists that killed the capitol police officer, ran sacked the capitol and almost disrupted a presidential election. many plotted under platforms. i think we understand violence is real. and so, this is why we're here today. in the jurisdiction with power to protect our fellow americans. mr. zuckerberg mentioned effective moderation systems. so, now you know you have systems to prevent these harms. thank you for championing support accountability. the question is what specific changes to section 230 do you support to insure more accountability? mr. zuckerberg just mentioned categories of content that are clearly illegal and privacy standards and votability and plm
puchai, should we be holding platforms accountable if they violate them under 230? >> congressman, first of all, there are many ways and many laws today which do hold us liable. fdcs oversight. hippa, areas where there are privacy laws and we have calls for federal privacy legislation. in california, we have privacy state legislation. we are both accountable, and subject to private action against these. efforts. is that correct? >> i definitely think what mark is talking about, along the lines of transparency and accountability, are good proposals to think through.
there are various legislative proposals. >> mr. dorsey, do you think we should have categories of content, u.s. privacy standards and probability, as well as standards for the violation. >> i believe, as we look upon 230, building upon it, i think we need more transparency around the content moderation practices. i think we need more robust processes and i think the real issue is evidence. and giving people more choice around algorithms. so, if there's any one i would pick, it would be that one. it's a tough one. but the most impactful. >> thank you, mr. dorsey. and misinformation went rampant in hispanic and florida when the presidential election. even with the politica admin,
how do you think this happened? mr. zuckerberg? >> congressman, it's -- i do think there's too much misinformation across all the media we've talked about today. how did it happen? we've talked a lot today about algorithms. i think a lot of this stuff happens in what we refer to as determinestic products. someone sends a text message. there's no algorithm. people can just send that to someone else. a lot of the stuff, unfortunately, was amplified in tv in traditional news as well. there was some of the content on facebook. and it's our responsibility to make sure we're building effective systems to reduce the spread on that. i think they performed well in the election cycle. there are always going to be new
things we will needed to do to keep up with the different threats we face. >> we've committed to reduce these to moderators and to keep this from happening again. >> congressman, this is something we focus on. we already beefed up and added more capacity to spanish-language fact checking and authoritative information resources. and that we're set to build on the future. >> the chair recognizes for five minutes. >> thank you, mr. chair and to the witnesses. i represent constituents in the great state of arizona. most of my constituents just want to be treated fairly, equitably, impartially, and they want to make sure that their private information stays private.
mr. puchai, did the influence google search results? >> we do index and wikipedia is in our index and for certain queries, if an answer from wikipedia rises to the top of our ranking, yes, we do rely upon it. >> did you personally decide to ban president from your platform? >> we have a process that we go through to get there and that came after a warning. >> and did you make the final decision? >> ultimately, i had final responsibility. >> thank you. and mr. puchai, the "wall street journal" reported that google lent hundreds of outside developers scan inboxes of millions of g mail users.
do google employees review and analyze g mail users' content? >> congresswoman, we take privacy very seriously. we don't use the data from g mail. and our employees generally do not access it. only narrow cases, either to trouble shoot with the right permissions. there are promissions. with enough checks and balances. >> i think what you're saying is occasionally your google employees review and analyze. i have another question regarding that. did google share g mail user's emails or analysis with third parties? >> we did not sell any data. i think what you're referring to is users could give a.p.i. access to third party. for example, there were
applications, which could give travel-related information. it's an a.i. on ton of the platforms. we have done numerous tests to make sure users have to go through steps before they get sent to a third party. >> and so, i've looked through your google privacy statements and user content and i still have concerns about that. i'm very concerned. i have g mail accounts, just like millions of people. and i don't know if you're looking at them, i don't know who is looking at them, who's sharing them, i don't know what you're doing with them. it makes me concerned. i -- >> one thing i said there. only if a user asks us to trouble shoot an account with the user permission. but we do not look into user's email contents and we do not share the contents with anyone else without the user's asking us to do so.
>> however, the "wall street journal" had this article saying that hundreds of developers were reviewing the email content. so, i have to move on to another question because i only have a short time. mr. dorsey, twitter denied the center for immigration studies the ability to promote four tweets that contained the phrases illegal alien, even though those are the correct legal terms. if there is a warning poegsed related to a boarder threat, how will twitter algorithms react to the word illegal, verses undocumented? >> it's not about our algorithms. it's interpretation against our policy but we can follow up on you about how we handle situations like that. >> this is the legal term is illegal alien. that is in law. a legal term. so, i don't understand why you would not allow that.
that is the legal, factual term. and with that, i'm going to ask another question. mr. zuckerberg, this has been brought up before. do you believe that your platform -- >> this is something we studied and care a lot about. products that improve people's well being is important to us and it helps people stay connected, which is one of the most fundamental and important things that we do, whether it's for teams, or for people that are older than that. and again, our policies on the main apps that we offer generally prohibit people under the age of 13 from using the services. >> general lady's time is expired.
>> chairman, i am enlightened. by the panel today. the three of the most knowledgeable business people in the world, and business models, sense of a future direction your companies want to go in. standards that are, in many cases liable. and a very big concern by the congress of the united states. and the direction you want to go in. verses what's gofor our nation in total. mr. zuckerberg, last october, 54 pages and 76 instagram accounts for violating your inappropriate
behavior policy. you know, you forge network was based in arizona. and ran its disinformation operation from 2018 to 2020. by creating fake accounts and commenting on our other people's content about the 2018 midterm election, the 2020 presidential election, covid-19, and criticism for, and praise of creation. certain political parties and presidential candidates. sadly, facebook only acted after a "washington post" investigation reported its findings. your testimony states 2017, facebook has removed over 100 networks of accounts for engaging in coordinated
behavior. where did facebook fail by not finding this network over the course of a number of years? >> well, congressman, we have a team of, i think it's more than 300 people that work on counterterrorism at this point and basically, trying to work with law enforcement and across the industry to find these networks of fake accounts and authentic accounts trying to spread behavior. i think we've got an lot more effective at this. i can't say we catch every single one. but certainly i think we've gotten a lot more effective. including this week we announce we've taken down a network of chinese hackers outside of china. so, we've got more sophisticated at this. sometimes when we start finding
a lead, we need to wait to see the full extent of the network so we can take down the whole network. so, that's a trade off we're able to discuss with law enforcement and other times not. in terms of how we do enforcement. i think this effort has got a lot more sophisticated. >> the amount of personnel you have working on these issues? >> i think we have one of the leading teams in this area. >> are you -- the question was are you happy with the amount of people you have in the capacity you have in these issues? >> the team is well staffed. well funded. billions a year on this content and security issues across the company. so, i think that is appropriate to meet the challenge. there are always things we're going to want to do to improve the tactics and a lot of it has been increasing the work that we
do with law enforcement and the intelligence community. >> let me ask you another question, mr. zuckerberg. thank you very much. i do want to say that, again, you're a bright, intelligent ceo. you know in advance what you want. your algorithms are created by your company and the other companies. you have control over the algorithms. so, the idea that you're in this direction, mr. zuckerberg the most recent community standards states that 2.5 million content related to suicide were removed in the fourth quarter of 2020. you can do this if you want to do it. very briefly, what policies facebook put in place to increase viewer capacity. not just on that issue but how
much over time can you review the capacity? >> sure, congressman. the biggest thing that we've said and more than 95% of the height speech that we take down is done by an a.i. and not a person. i think it's 98 or 99% of the terrorist content is identified by an a.i. and not a person. you mentioned the suicide content as well. i think high 90s percent is by a.i. -- >> i'm out of time. i want to state very briefly. you have a lot of work to do, you and your cohorts on this panel. thank you. >> chair recognizes mr. pence for five minutes. >> thank you chairs, doyle and ranking members for holding this
joint subcommittee hearing. and thank you and the witnesses for appearing before us. the extent to which you're rem nessant to the all encompassing entity we've seen in the past century. in the early 1900s, standard oil had a monopoly over our country's supply. in the 1970s, we knew the telephone was going to be ma bell systems. in each instant, participation in society. in a similar sense t is difficult, if not impossible, to participate in society today without coming across the platform. you can choose not to. but it's considered essential.
even the government has become an equal contributor. each member of congress and every senator is all but required to use your platform to communicate with their constituents while they're in washington d.c. i know you understand your platforms have a responsibility to act in good faith. and unfortunately, regularly the face on twitter account and many of my peers and are littered with hateful and between constituents and in the idea. your platforms claimed that you've referenced today. i'm your sure your aware government accounts have a limit the ability to maintain a
platform with the productive resource of information for the public. they have essentially become a microtown hall without a moderator on social media. i agree with all your testimony that a deficit has been growing over the past seven years. to adjust it, we need to do something about it now. the way you manage your platforms in an inconsistent manner has deepened the mistrust and divide the public contraversion. and they've told me they are increasingly mistrustful of platforms that enforce your policies. there's just a few examples of how this is occurring. members of the chinese communist party have verified twitter accounts to regularly pedal all misleading claims surrounding
the human rights violations we know are occurring in northern china. twitter gets the supreme leader of iran a mega phone to proclaim derogatory statements endorsing violence against u.s. and western culture. twitter accounts associated with the supreme leader, that call israel a cancerous tumor, and call for the ired aication of the zionest regime. this, as he bans the service for his own people to restrict their expression. clearly you need to do more to address content that violates your policies. i have two questions. why is the chinese communist party allowed to continue with your platform after propaganda to cover up human right abuses against muslims in northern china, and two, why does the
supreme leader of iran still have a platform to make threats against israel in america? >> so, first and foremost, we do label those chinese accounts and people have context as to where they're coming from. a tweet so people understand the source. we think that's important. we are reviewing our world leaders policy. we're taking public content for review right now. we're enabling anyone to give us feedback. >> if i may interrupt you quickly on that very point. iran has been supporting hezbollah and it's not just sabre rattling, as you've made the statement, or your company's made the statement. they have some serious damage to hold countries and people. and as i served in the military,
they killed hundreds of marines many years ago. so, i don't know what you to have study about this. mr. chairman, i yield back. >> gentleman yields back. chair recognizes ms. rice for five minutes. >> what is winning, yes or no, on your twitter account poll? >> yes. >> your multitasking skills are quite impressive. in december 2020, the house committee released a report entitled hijacking our heroes, exploiting misinformation on social media. i ask that this report be submitted for the record. >> so ordered. >> i heard of the report because it's deeply disturbing the involvement of our veteran and military service members and the violence that took place january 6th. it's estimated that one in five people are serving or were
served in the u.s. military. it should come as no surprise, that for years nefarious actors have used to harness algorithms to impose harm on service members that they did not seek out for themselves. they're particularly targeted by malicious actors online to misproepriate their voices. we have to do better for those who have served our country. do you believe that veterans hold a special status in our communities and have military training, making them prime targets for domestic terrorists and our adversaries seeking to form an insurrection? >> congresswoman, i certainly believe veterans hold a special place in our society. i haven't -- >> did you see at the capitol and nation ltal mall, there were rioters armed with tactical
equipment? did you see those images? yes or no? >> yes. >> have you personally talk toads iraq, afghanistan, veterans of america about disinformation campaigns targeting veterans? >> no, i have not personally. although, our team is in content with a number of these groups, as we set up our policies. >> have you talked to the vietnam veterans of america about disinformation campaigns? >> congresswoman, i can get back to you on whether our team has consulted with them specifically but -- >> thank you. do you believe veteran and military service members are just like other marecons, in that they're susceptible to impulses in human psychology that facebook exploits in human psychology? do you believe they're susesable in that way? >> there's a lot in their characterization i disagree with. erse
>> do you think they're susceptible to that kind of information coming at them? yes or no. >> congresswoman -- >> given your answers, i'm not confident you have the resources needed to eradicate misinformation. would you support legislation that require you to create an office of veteran's affairs that reports to the ceo and worked with outside veteran service organizations to insure our enemies don't gain ground trying to radicalize our brave men and women who served in our military. would you support that legislation? >> congress wmp, i think details mat arlot. so, i would be happy to follow up with you or have your team follow up with our team to discuss this. >> it's just a broad stroke. do you believe that you could
find your way to support legislation that would have, the protection of our military, active duty, and veterans? >> i think something like that certainly makes sense. >> so, i devote to you, mr. zuckerberg and a misinformation campaign. i'm curious if you're familiar with the word veteran or public pages did you remove from your platform after january 6th in association with misinformation about the 2020 election or the attack on the capitol? >> i don't know the answer off the top of my head but i'd be happy to get back to you. >> your platform was, in fact, a crime scene after january 6th. we need that information and data to understand how that
attack happened. thank you for spending so much time with us. i yield back. >> gentle lady yields back. recognizes mr. armstrong for several minutes. you need to unmute. >> all right. sorry about that. can you hear me? >> yes, i can hear you. >> all right. thank you. no other industry receives such bipartisan scrutiny. deplatforming, antitrust and the list continues to grow. we discuss these things to often in isolation but they're all related. it starts with the fact that the users are not your customers, they're the product. more specifically, the data you collect is your product.
you're to use the data for behavior advertising. data's unique as a business asset. it's perpetual and reinforcing. data begets more data. massive data collection harms competition. that's why censorship is so concerning to all of us. your platforms have a strangle hold on the flow of modern communication and i think we have to resist the urge of content moderation and censorship. in 1927, justice wrote, the remedy to apply more speech, not enforce silence. i think that statement still holds true today. yet, your platforms don't simply silent certain speech, the algorithms are forced to design because you profit by keeping users locked into ininformation
. at that hearing i raised several examples of the ad tech staff. your answers largely reiterated privacy justifications which i understand and support. however, my question was whether google's consolidation of digital advertising would further harm competition? since then, i would view the privacy sand box and privacy proposal, which is to replace third-party acteders. this is my question. how will these actions not furgter in -- further intrench google's competition? >> as you rightly point out, privacy is really important. users are giving clear feedback in terms of the trend they would like to take.
advertising allows us to provide services to many who would otherwise not be able to use services. we're trying to provide relevant ads, protecting their privacy. that's what flock is working on. >> i going to move on because i understand the privacy. i understand the privacy. and i understand the rational of eliminating individual-level tracking and the potential privacy benefits of user data in chrome at the device level. but this is still eliminating competitors' access to user data at a time when you already control 60% of the market. and not actually benefit user privacy n stead of spreading it, it will be with you. and so, i guess my point is congress needs conduct careful oversight, as they're
introduced. and we need to insure that competition is not stifled further. i have one question and it's important. i'm going to ask all three of you. when we're conducting competition analysis in the tech industry, should nonprice factors like privacy be considered? >> i think so. i think privacy is very important. we have called for comprehensive privacy legislation. google doesn't get any access to flock data. it's protected. and then we will publish more papers on it. >> i understand completely. i mean, you're forcing advertisers into the ad staff. i don't discount it. it increases privacy. i think this is a real problem because i think they're in conflict, with each other. do you think when we're conducting analysis, non-price
factors should be considered? >> not sure exactly. open to further discussion on it. >> how about you, mr. zuckerberg? >> yes, congressman and my understanding is the law already includes quality to product, including price. >> i appreciate you talking about big platforms and small platforms. i think in our history congress has done a good job and with my last six seconds. please all dado a better job making sure artists get paid better on your platform. >> thank you, mr. chairman. it's often said that lies travel faster than truth and we've seen that play out with devastating consequences today.
this concerns me greatly, not just as a father or lawmaker, but as someone ready to see the past divisions that have dominated our country for the past several years and really decades, really. but it's hard to see how this can change, when the ceos of the largest social media companies repeatedly say they've fixed their ways, only to keep spreading harmful lies and misinformation. last august, here in the dallas, fort worth area, the north texas poison control center felt the need to warn people against ingesting bleach or other disinfectant products as a cure to prevent covid-19. despite efforts of your companies to take down such harmful misor disinformation calls to the north texas police and control center about disinfectant ingestion rate said were higher than usual.
and state-wide calls about bleach products were up over 70% compared to the year before. the north texas poison center pointed out largely to misinformation online as the cause for these increases. and as we know, in the lead up to last elections, black communities were specifically targeted for disinformation campaigns designed to suppress the vote, especially in battleground states and right now there are sites up that are discouraging people from getting the covid-19 vaccination. i know a lady who was put in jail for 30 days, because all she did was repost a faulty post that said black folk aren't falling for this b.s. and she was put in facebook jail for 30 days. now, if they're taking down or labelled false, lies travel a lot faster than truths.
your companies have been largely flat footed when it comes to getting out ahead of these issues and it's time for something to change. that's why i'm exploring legislation that would establish an independent organization of researchers and computer scientists in this trend before they become viral. this early warning system will is disinformation is spreading. mr. zuckerberg, would you support legislation of harmful disinformation and conspiracy theories spreading across your platforms? >> congressman, i think we need to look into that in more detail to understand the nuances. in general, i agree to slow the spread of this kind of misinformation.
and that's why we've taken all the steps outlined today, from building an unprecedented independent fact checking program too, taking down content that could cause imminent physical harm,to the work of the covid information center and the climate information center, to promote authoritative information across our services. i think there's a lot to do here. >> mr. dorsey, would you support legislation? >> it will be very much whac-a-mole. i think the most important thing is to provide data everyone has access to and review. >> i have a 14-year-old at home that watches youtube. what about you for those
platforms? >> already today, in many of these areas, we show, proactively information panels. on we shared a lot of information and we have use of 400 billion and active information, including information panels, i think makes sense to me. >> i appreciate the time, mr. chairman. i think we need to act quickly. we're running out of time. i yield back nigh time. thank you. >> the chair now yields five minutes. >> thank you so much, mr. chair. mr. zuckerberg, thank you so much for joining us today. as cochair of the lgbtq equality process in the u.s. congress, i'd like to ask you a few
questions about an incident that occurred several weeks ago now and i would appreciate a simple yes or no answer. most have no room for nuance. and facebook took down a video posted by my colleague, newman, in which she places the transgender flag outside her office. is that correct? to your knowledge? yes or no? >> i'm not aware of this. >> you're not aware of this. >> the answer is yes. facebook took her video down. according to representative newman, the reason facebook gave for taking down the video was it violated facebook's community standards on hate speech in superiority. does that seem right to you you that if someone took a video and
posted it on your platform, that it should be took down? >> congresswoman, that doesn't seem right to me, but i would need to understand the specifics of the case in more detail. >> yes, thank you. the answer is no, it's absolutely not right. meanwhile, across the hall, representative margery taylor green from georgia posted a video to facebook, her video showed her putting up a transphobic sign so, that a mother of a trans could quote, see it every time she steps out her door. they allowed representative green to stay online? yes or no? >> as i've said a number of times today we do make mistakes, unfortunately and we hope to fix them as quickly as possible.
>> the answer was yes. representative green's video was allowed to remain online. representative newman reached out to facebook and a few hours later her video was restore would an apology but representative green's video was never taken down. i'm not even going to ask if i'm getting that right, as i was, because you obviously don't know. are you aware that facebook has repeatedly flagged the transgender flag of hate speech and that transpositive content ends up being taken down, while transphobic content is often shared widely. yes or no? >> congresswoman, i'm not aware of that specifically. this is in instance of a broader challenge in identifying hate speech, which is there is a very nuanced difference between someone saying something racist, verses saying something to
denounce something someone else said and we need build systems that handle this content in more than 150 languages around the world and do it quickly and unfortunately, there are mistakes in trying to -- trying to do this quickly and effectively. >> mr. zuckerberg, i'm going it give you your nuance this one time. as it exists today, do you think your company is going to get these content moderation decision right on the first try eventually? >> congresswoman, if what you're asking is are we ever going to be perfect, the answer is no. i think we'll get increasingly accurate over time. for example -- >> i only have a couple of minutes -- one minute left. so, i'm going to continue here. as it's been mentioned repeatedly throughout the day, we don't have faith that your companies have the proper incentives to proactively contemplate and address.
with that in mind, would you require legislation to allow social media companies reporting to the ceo. and that would mean you would have to reconsider your corporate structure, including civil rights and human rights of the transcommunity? >> congresswoman, we took the unprecedented step of hiring a v.p. of civil rights and i think we're one of the only companies that has done something similar to what you're saying. >> i hope that you do better then. because this example i'm giving you was completely unacceptable. this panel has done something truly rare in washington these days. er for it's united democrats and republicans. your industry cannot be trusted or regulated and with that, i yield back. >> the chair now recognizes miss transfor five minutes? >> i'd like to turn the focuses
back to our children? my husband skb i have five. the oldest is 27, the youngest is six and over the years i noticed how technology has been increasingly designed to capture their attention. face interactions with us. keeping our children off of 13 back to you. you were actively on boarding our children on to your ecosystems with apps like youtube kids, facebook messenger kids and now we're hearing instagram for kids. these applications introduce our children to social media far too early and include manipulative design features intended to keep them close. when a child finish as video, does the next video automatically play by default?
and i think this is a yes or no? >> aeroy. sorry. i was muted. i have children too. i worry about the time they spend online and agree it's an issue. >> the autoplay function by default is a yes? >> on the main app, it's there. users have -- >> the default setting is yes. when a user, who was predicted to be a teen is watching a youtube video, are the number of likes displayed by default? yes or no, please? >> on all videos i think we do have across all videos we have. >> great, and mr. zuckerberg, will the recently reported instagram app for kids have endless scroll enabled? yes or no? >> sorry. congresswoman, we're not done finalizing what the app is going
to be. we're pretty early in designing this. >> are you not sure or not sharing features? or -- and look, another fiche of concern is the filter, that, as an unnatural or crooked flow from 10-year-olds who apply to her face. is that feature part of instagram for her kids? >> i haven't discussed this with the team yet. >> look, please expect my office and many others to follow up, given what we know about instagram's impact on teen mental health, we're concerned about our younger children. and just -- i just want to speak mugter to father for a moment because leading experts all acknowledge that social media sites pose risks to young people. inappropriate content, over sharing of personal information, cyber bullying and those risks
are exacerbated with more time children spend in these apps. you mention that you have children. and i've also read you limit their screen time. what do you say when one of the children doesn't want to put their phone down? >> congresswoman, the struggle is the same, particularly through covid. it's been hard to moderate it. i do take advantage of the parental controls and digital well being tools. we have provisions in place. >> i don't mean to cut you off, but the last thing over worked parents need right now, especially right now, are more complex to dos, which is what parental controls are. i mean, they need child-centric design by default. mr. zuckerberg, i understand your children are younger, but when they start using social media, what wim you say when they're craving the tablet over
face-to-face time with you or their friends? >> we're designing the tools. we design messenger kids that the parents are in control. that was one of the things that made us think we should consider doing this for instagram as well, by having it where we have a parent-controlled experience and as you -- >> i am going to reclaim my time, only because conecing with others is one thing. adding filters, no breaks for kids to take, and manipulating the design of these apps for our children is another. look, this committee is ready to legislate to protect our children from your ambition. what we're having a hard time reconciling is that, while you're publicly calling for legislation, which comes off as incredibly decent and noble,
you're plotting your next frontier of growth, which targets our young children and you take great strides, with infinitely more resources in protecting your own children. this play book is familiar. it's the same tactic we saw from alcohol companies and big tobacco. start them young and bank on them never leaving or at least never being able to. but these are our children and their health and well being deserve to take priority over your profit. >> gentle lady's time is expired. chair now recognizes miss flecher for five minutes. >> thanks to you and ranking members. i agree with my colleagues, there's a broad concensus on a range of issues and i appreciate the discussion. as we've discussed extensively, one of the big challenges is it denies us a basic set of shared
backs to enable informed debate, like what we are having here today. and it's absolutely vital that we take charge and that we address it. what we've seen is country's interests are not aligned with ours. extremists organizations have used platforms to amplify from the covid-19 pandemic to the insurrection. when we see the real-world cost of misinformation and i worry the structure of many social media companies, including those before us today, prioritize engage; including with extremists content over citizenship. one of my greatest concerns regarding how extremists content and misinformation is allowed to spread on the platform is a lack
of data transparency, when it comes to independent analysis. anyone with an internal system, it's about the system. but we have no way to verify how effectival those systems are. and that's a huge part of the challenge before us. i think we all would agree that we need data and information to make good policy and right legislation that will be coming out of this committee. that brings me a follow up with colleagues. it's my understanding that all three of your platforms chose to remove content that was posted regarding the insurrection january 6th. as a result, it's unavailable to researchers and congress. my question is will you commit to sharing the content to inform our investigation of the events of january 6th and the issues before us today, about how to
respond to extremists and dangerous content online? and i'll start with mr. zuckerberg. >> thanks, congresswoman. when we take down content that might be connected to a crime, i think, as a standard practice, try to maintain that so we can share with law enforcement, if necessary. i'm sure our team can follow up to discuss that as well. >> i understand you have a legal obligation to cooperate with authorities and law enforcement in these cases. what i'm talking about is with sharing with us in congress. i appreciate your response there. mr. dorsey. >> we've been thinking about a roprogram to get researchers to actions we take but all of this is subject to local laws, of course. >> and that may be something we can help craft here. something we've herds from researchers as well is a real area of challenge in not having
the data. and mr. pachai? do you also agree? >> congresswoman, sorry. was muted. we are working with law enforcement and happy to connect with your office. while balancing the privacy of the people. >> thank you. so, i appreciate all of your willingness to work with us and assist congress in addressing this attack on our capitol and country. another idea i would like to touch base with you on, in the time i have left, just over a minute, is the difference we see in how your platforms handle extremists content, verses citizen content? and they do a better job from isis or al qaeda and others, where the posts are automatically removed, depending on key words and phrases, etc. but ftos are designated to
identify groups that could cause harm to americans. there's no legal mechanism or definition for doing the same for domestic terror and hate groups. would a federal stand frrd defining a domestic terror organization better help remove hateful content from your site? mr. zuckerberg? >> i am not sure. domestically, we do classify a number of white supremacist organizations and militias and conspiracy networks like qanon as the same level of problematic. -- >> i hate to cut you off. your answer was i'm not sure. a quick yes or no from mr. dorsey and mr. pichai. >> very quickly because your time is expired.
>> we need to evaluate it. >> focus on it, i think we're happy to work and cooperate. >> thank you. i yield back. >> it's my understanding we have -- let's see. eight members who are requesting the wave on for the hearing. i believe we have given all members of the subcommittees their opportunity to speak. so, we'll now start to recognize the members waving on and first on the list, i see mr. burjs. are you with us? >> yeah, i am. thank you, mr. chairman and thanks to our witnesses for spending so much time with us. and this is clearly a very important issue to every member of this committee, regardless which political party they identify with. let me ask you a question because it strikes me, listening
to your answers to both the colleague, jeff duncan and angie craig, both coming at the issue from different directions. but the concern is that there was a -- there was the exercise of editorial authority over the postings that were made on your website. is that a fair assessment? >> congressman, i'm not sure what you mean but i think content moderation and enforcing standards, i don't thik that's the same judgment that, for example, a newspaper makes when writing a post? >> but maybe it is. mr. duncan eloquently pointed out there was restriction of conservative speech and colleague craig pointed out how there was restriction of trans affirm speech.
that seems we're close to exercising editorial discretion. forgive me for thinking that way. and i'm sure i'm not alone in this. it does call into question the immunity provided under section 230. maybe it is not a problem with the law itself, section 230. maybe the problem is that the missions has changed in your organization and other organizations. >> congressman, i'm not sure what you mean, but we have clear standards against things like child exploitation, incitement of violence, pornography, things that -- >> i'm -- all spelled out in plain language of section 230. but again, you're putting restrictions on conservative speech, mr. duncan pointed out how that is occurring.
and craig pointed out restrictions on trans-affirming speech, none of those fall in the other categories you're describing. so, to the casual observer, it appears you're exercising editorial authority and maybe you should beregulated as a publisher, as opposed to someone indifferent to the content they're carrying. >> congressman, i think one of the virtus of section 230 is it allows them to moderate things like bullying that are not always illegal content but you and i would agree are harmful and bad. so, i think it's important that companies have the ability to go beyond what is legally required. i do not think that makes these internet platforms the same thing as a news publisher, who is literally writing the content themselves. i think we have more
responsibility than maybe a telephone network. >> let me interrupt you with interest of time. mr. dorsey, every presidential tweet i read, following the election had an editorial disclaimer appended to it by you. how does that not make you someone who's exercising editorial discretion on the content that you're carrying? >> our goal with our labels was simply to provide connection to other data and provide context. >> yeah, but you don't do that routinely with other tweets. it's a singular assignment someone had taken on to look at whatever the president is publishing. we're going to put our own spin on that. and again, that strikes me as an editorial exercise. the only reason i bring this up and we're going to have these
discussions. i recognize smaller companies just starting out, protection of section 230 may be invaluable to them. you're established, mature companies. you're mature companies. you exercise enormous control of the thought processes of not only the entire country, but the tire world. i think we need to revisit section 230 in the terms of have you now become actual publishers as opposed to simply carriers of information. thank you, mr. chairman. i yield back. >> the chair recognizes mr. tomko for five minutes. >> thank you. gentlemen, thank you for being with us today. there are many issues i would like to raise with you. the mes pressing unresolved questions reinvolve around january 6th when i had to dive for cover in the house gallery
as violent insurrectionists attempted to break down the doors and take the chamber. they were pro he peld by one believe, that the election had been stolen from former president donald trump. they reached this false and dangerous conclusion without evidence yet somehow in massive numbers. their assault was not disorganized or isolated and not quince kens. mr. zuckerberg you have downplayed the role facebook played in helping the rioters mobilize on january 6th. in light of growing evidence that suggests otherwise, including the fact that facebook was the most cited social media site in charging documents, the department of justice filed against insurrectionists, do you still dough my your platform was used as a significant megaphone for the lies that fueled the insurrection? >> congressman, to be clear, i think part of the reason why we're -- why our services are cited in the charging docs is because we worked closely with
law enforcement to help identify the people who were there. so i don't view that collaboration with law enforcement should be seen as a negative reflection on our services. and as i have said a number of times today, there was content on our services from some of these folks. i think that was problematic. but by and large, i also think that by putting in place policies banning qanon, banning militias, banning other conspiracy networks we made our services inhospitable to a lot of these folks and that had the unfortunate consequence of having those folks not use facebook and use other places as well. so there is certainly more for us to do, but i stand behind the work we have done with law enforcement on this and the systems that we have in place. >> thank you. mr. pichai, can you state that youtube did not recommend videos with stop is the steal content,
white supremacy content and other hate and conspiracy content that were seen by rioters at the capitol? >> congressman, we have clear policies and we were vigorously enforcing. leading up to the election we removed hundreds of thousands of videos and terminated 8,000 channels, and on the day of the riot we were successfully able to take down inappropriate live streams. we gave precedence to journalist journalist i can precedence. since then we have been cooperating with law enforcement as well. >> so you are indicating that you did not recommend videos with stop the steal? >> we were rigorously enforcing clear policies around content that undermined election integrity. once the states certified their
the election, we take down videos which were violative. so we have been monitoring it very closely. >> thank you. mr. dorsey, are you confident that the conspiracy theorists or other purveyors of electoral disinformation and stop the steal on twitter were not recommended to others? >> i can't say that with confidence, but i know we worked hard to make sure if we saw any campfication that went against the terms of service which would -- we took action immediately. we didn't have any upfront indication this would happen. so we had react to it quite quickly. >> thank you. so who and what content do your platforms recommend, have real world con zwenss and the riot caused five deaths and shook our democratic foundations. i believe your platforms are responsible for the content you promote and look forward to working with my colleagues to determine how to hold you
accountable. mr. pichai and google and youtube often slip under the radar as a source of disinformation, but in the last election bad actors used ads on google search to scam people looking for voting information and youtube failed to remove videos that spread misinformation about the 2020 vote results. so, when generalists pointed out in november that election misinformation was rampant on google's youtube the company said it was allowing discussions of election processes and results. a month later it said it would remove content alleging widespread voter fraud in the election. why did youtube take a month? >> if i could clarify here, we were taking down videos leading up to the election. we did -- there is, obviously, a month from the date of election to the due processes, challenges and we concerted with the association of secretaries of
state and on december 8th when the states certified the election, we started enforcing newer policies on december 9. to be very clear, we were showing information from "the associated press" and we were proactively showing information high up in the search results to give relevant information throughout this election cycle. >> the gentleman's time is reco mckinley for five minutes. >> this panel, you all have to be exhausted after being grilled all day long like this. my questions are to mr. zuckerberg. when you came before our committee in 2018 you acknowledged that facebook had used what you just said clear standards preventing the sale of illegal drugs on your site, but you were shown examples of active -- that traffickers were using that platform unlawfully
to sell prescription opioids. now, you did apologize and confirmed that, quote, social media companies need to do a better job of policing these posts, closed quote. now, three years later, it appears a show game is emerging. facebook seems to have cleaned up its act, but you are now allowing instagram, one of your subsidiaries to become the new vehicle even though instagram has the same policies, you are still allowing bad actors to push pills on your site. it didn't take long for our staff to find numerous examples. for example, here is oxycodone that is being sold on your site. here is ritalin being sold on your site. here is xanax and add roll that is being sold on your site. so those posts are not new. they have been active since last fall. if we can find posts this
easily, shame on you for not finding them for yourself. apparently, you are not taking the warnings of congress seriously. after drug manufacturers zufrmd millions of pills into our community killing thousands, ravaging families and destroying livelihoods, congress responded by passing laws to hold them liable. if a retail store is selling cigarettes to underage kids, that store is held liable. so why shouldn't you be held liable as well? do you think you are above the law? you are knowingly allowing this poison to be sold on your platform into our communities, to our children, to our vulnerable adults. look, i read scott galloway's book. i encourage all the members on the committee, it's a perfect depiction of arrogance of big
tech companies like google, facebook and amazon. he develops a very compelling argument as to why big tech companies should be broken into smaller companies much like that occurred in at&t in 1984. maybe it's time for congress to have an adult conversation about this loss of liability protection and the need to reform our antitrust laws. i don't think congress wants to tell you how to run your company, but maybe it should. so, mr. zuckerberg, let me close this one question. and so don't you think you'd find a way to stop these illegal versales on your platforms if you were held personally liable? >> we don't want this content on our platforms. i agree this is a huge issue. we have devoted a lot of
resources and built systems that are largely quite effective at finding and removing the content, but i just think we what we need to understand is the scale these communities operate where people are sharing millions or in messages billions of things a day, it is inevitable we will not find everything, just like a police force in a city will not stop every single crime. >> i agree. i asked you the question very directly, mark. should you not be held liable when people are dying because your people are allowing these versales to take place? we did it with manufacturers. we do it to the stores. why aren't we doing it to the salesmen that allows this to take place? >> congressman, i don't think we are allowing it. we are building systems that take the vast majority of the content off the symptoms. >> three years this has been going on and you said you would take care of it last time. you switch from facebook over to
instagram. they are still doing it now. you are saying we need to do more. how many more families are going to die? how many more children are going to be addicted by you still studying the problem? i think you need to be held liable. >> congressman, we are not studying the problem. we are building infect systems that work across facebook and instagram. i am saying i don't think we can expect every platform will find every instance the harmful content. we should hold the platforms responsible for building generally effective systems at moderating this types of content. >> the gentleman's time is done. >> thank you. >> the gentleman yields back. the chair recognizes miss rock esther for five minutes. >> thank you for allowing me to weigh into on this important hearing and thank you to the witnesses. i want to focus on two areas. first, a consumer protection and safety issue, and second, more
broadly, manipulation and privacy of our data. on consumer protection and safety, earlier this year two infants from two different families ended up in the intensive care unit in wilmington, delaware, ever being fed homemade baby formula based on an instructional video viewed on youtube. one infant suffered from cardiac arrest that resulted in brain damage. for years the american academy of pediatrics has warned parents against homemade baby formulas because it puts infants at risk of serious illness and death. since at least 2018, the fda has recommended against the use of homemade formula. even as recent as 29 days ago the fda issued an advisory against it. in february, my office informed your team, mr. pichai, and as a follow-up, i have sent a letter requesting information and action on this issue and hope in the hopes of a response by april 1st. mr. pichai, this is just a
yes-or-no question. can i count on a response to my letter by the deadline of april 1st? >> congresswoman, definitely, yes. heartbreaking to hear the stories. we have clear policies. thanks for your highlighting this area. i think the videos have been taken down and we are happy to follow up and update with you. >> we checked today. for years these videos have clearly violated your own stated policy of banning the videos that endanger, as you say, quote, physical well being of minors. and so i'm pleased to hear that we will be hearing back from you. while we are considering section 230, what's clear is we should all be concerned by all of your anlts to adequately and just as importantly rapidly moderate content. and some of these cases we're talking life and death. second, as many of my colleagues have noted, your companies profit when users fall down the rabbit hole of disinformation. the spread of disinformation is
an issue we all grapple with across the political spectrum. it finds its way to the people most susceptible to it because the profiles that you create through massive data collection suggest what they will be receptive to. i introduced the detour act to address common tactics that you used to get such personal data as possible. and these tactics are often called dark patterns, and they are intentionally deceptive user interfaces that trick people into handing over their data. for the people at home, many much you may know this as when you go on an app that doesn't allow you to have a no option or insinuate that you need to do something else, install another program like facebook messenger app to get on facebook, you collect and use this information. mr. pichai, yes or no, would you oppose legislation that banned the use of intentionally
manipulative design techniques that trick users into giving up their personal information? >> we definitely are happy to oversight on these areas and -- >> thank you. mr. dorsey, yes or no? >> open to it. >> mr. zuckerberg? >> congresswoman, i think the princele am -- >> yes or no, please. >> your company recently conducted this massive ad campaign on how far the internet has come in the last 25 years. great ad. you ended with the statement, we support updated internet regulations to address today's challenges. unfortunately, the proposal that you direct your viewers to fails to dark patterns or deceptive design choices. mr. zuckerberg, will you have that as part of your platform for enter internet regulations? >> congresswoman, i'll think about it. my initial response is that i
think that there are other areas that i think might be more urgently in need. >> that might be your -- if you say this is a desire of yours to address the issues that we face today, dark patterns goes back to, you know, 2010, you know, this whole issue of deceptive practices. i hope that you will look into it. i will say miss trahan and others have mentioned, she mentioned our children, seniors, veterans, people of color, even our very democracy is at stake here. we must act and assure you we will, we will assure you we will act. thank you so much. and mr. chairman, i yield back six seconds. >> thank you. and now the chair recognizes mr. -- for five minutes. >> according to new data from the national center for miss ng and exploited children, found that the vast majority of child
exploitation reports were on big tech sites. facebook had the most. 20.3 million. google was second with 546,000 plus. twitter had 65,000 plus. put in perspective, mind you, the canada-based parent company of major porn sites had 13,229. facebook had 90% of the flagged incidents were duplicates. let's accept that. that still leaves over 2 million incidents. 2 million incidents. mr. zuckerberg, yes or no, does facebook have a problem with child exploitation on its platform? >> congressman, this is an area that we work on a lot. the reason why those numbers are so high is because we are so proactive about trying to find this and send to others who are doing good work in this area.
we send content and flags over to them quite liberally whenever we think we might see something is an issue, and that's, i think, what the public should want us to do. not criticize us for sending over a large number of flags, but to encourage the companies to -- >> are you saying that you all have a problem and this is one way you are trying to work on it. mr. pichai, yes or no, do you agree with mr. zuckerberg that you all have a problem? >> congressman, sorry, i was muted. this is an area which we invest very heavily. we have been praised by several authorities. we work proactively -- >> so the answer is yes. mr. dorsey, yes or no, do you agree? >> if we see any problems, we try to resolve them as quickly as possible. >> you do have problems. the problem is when you are
talking about millions of incidents, and we take 90% of them as duplicates from the facebook data, millions of incidents where our children are being exploited with child porn on you all's sites. we have got to do better. i think for everything that we talked about today an independent industrywide review team like electronic industry did with the underright of the writers laboratory 150 years ago. i told y'all that when you were here before. nobody's done anything. i don't think it needs to be within your company. i think it needs to be outside. and in that vain i would say to google, special permission was given to moonshot cde to target ads against extremists key words. moonshot then directed thousands of individuals who searched for violent content to videos and posts of a convicted felon who
espouses anti-law enforcement, anti-semitic, and anarchist viewpoints. mr. pichai, are you aware of this problem? >> congressman, i am not aware of the specific issue. last year we blocked over 3.1 billion bad ads. 6,000 ads per minute. so we enforce vigorously. but i'm happy to look into the specific issue and follow up back with you. >> here's what happens. you partnered with an outside group that didn't do their job. what are your standards when you partner with an outside group? what are your standards and fill os fees because they sent people looking for violence to a convicted felon with anarchist and anti-semitic views. >> there is no place for hate speech. i am disappointed to hear of this. we will definitely look into it and follow back with you. >> well, i appreciate that. i recognize that.
but i had the same concerns that mr. mckinley had and you weren't here last time, but we heard these same kinds of things how we are going to work on it and fet these problems resolved and i forget whether that hearing was, but a year are so ago, yet we continue to have the same problems where political candidates information is being taken down because, for some reason, it's flagged, where conservatives and people on the left are being hit and taken down, and i free with many of the sentiments on both sides of the aisle that if you all are aren't doing thing, and it appears you are not moving fast enough, we have no choice in congress because but to take action. i don't want to. i'd rather see you all do it like electric industry did with underwriters laboratory. but nobody is doing that. nobody is coming up with a group that both sides of the aisle and the american families can feel comfortable with. and so we're going to have to take action and it's probably
going to be this year. i yield back. >> gentleman yields back. the chair recognizes miss schrier for five minutes. >> thank you, mr. chairman. i am a pediatrician and i have spent my life calming patients to are nervous about vaccines because of online misinformation. in fact, that's why i introduced the vaccines act when i was a new member of congress. did you know that there are doctors who, ar spending their entire day on the front line fighting this virus, they come home at night and spend their scarce free time and family time fighting misinformation of vaccines online. and this misinformation, of course, comes primarily from facebook and twitter. so why do they do that? well, they do it because of things like this that happen after i introduced the vaccines
act. here are some overt threats. keep shoving this vaccine monitor down people's throats and expect riots. be careful, you will answer for this tyranny one day. she needs to just disappear. can we vote her out of office? i'm enraged over these poison pushers. we have weapons and are trained to fight off possible forced vaccinations. i will die protecting my family. and then there is just the misinformation. it says safe and effective many times, yet no vaccine has been studied in a double-blind study. false. who is going to take this vaccine? i heard rumors that it changes a person's dna. false. you do not give -- excuse my language, a shit about the health and welfare of our children. you are deplorable and of course that again is false. so while the overt threats are unsettling, particularly after
january 6th i think about this whole ecosystem, yuriko system that directs a hostile sliver of society en masse to my official facebook page. these are not my constituents. most came from two specific groups that directed their members to my page. mr. zuckerberg, i have some questions for you. i know you understand these issues are important and sometimes misinformation can be very hard to spot. would you agree? >> congresswoman, i agree with both of those. this is important, and enforcement processes can be difficult. >> thank you. and i heard your answer earlier to representative upton's question that there are 35,000 people doing content review of posts that have been flagged by users in a.i. can you tell me what content review means and how many of those 35,000 are dedicated to topics regarding health? >> congresswoman, yes. with the people who are doing, what they are doing is content
gets flagged by the a.i. systems or by another person in the community and if the a.i. can't by itself determine that something violates or doesn't, then it gets flagged for further review and human judgment and 35,000 people go through the different queues, focused on the different harms we have discussed today. i don't have the number off the top of my head about how many are focused on vaccine misinformation. as you know, we have a policy that stunt allow vaccine misinformation and we work with the w.h.o. and cdc to take down false claims around covid and the vaccines around that, that that could cause harm. >> that's where it gets tricky because you have to have experts in health care professions who really understand. are your people trained in health care to really even be able to discern what's rail, what's fake, and what to take down? >> congresswoman, the people who set the policies either are experts in these areas or engage in a consultative process where they talk to these folks.
in this case we largely defer to the cdc and w.h.o. on which claims they think are going to be harmful and then we try to break that down into kind of very simple protocols that for 35,000 people can follow and that we can build into a.i. systems to go find as much of that content proactively as possible without requiring all those people to be medical experts. >> with my short time remaining, i would love to jump to that part about the cdc because i want to turn my attention to the covid resource center that you described as a central part of your efforts to fight misinformation directed over 2 billion people to the covid-19 information center, but on that information page almost all of the content links to additional facebook pages. it looks to me like an extension of facebook's walled garden that keeps users on the site instead of directly to authoritative trusted sources like the cdc.
knowing your platform is a large source misinformation did you refer directing people to the cdc rather than people them within your platform? >> congresswoman, i think we have considered both and we have done both in different cases. the team is very focused on building this in a way that will be most effective at getting people to see the content. and i believe they have concluded that showing content from people who within a person's community that they are going to trust on the services is one of the most effective things that we can do. >> general lady's time is expired. >> thank you. >> the chair now recognizes mr. crenshaw for five minutes. >> thank you, mr. chairman. thank you all for being here. it's been a long one. i have been on social media longer than anyone in congress, i think. i was one of the first schools to have facebook back in 2004.
it seemed that the goal of social media was simply to connect people. the reason we are here today is because over time the role of social media has ex packed in an extraordinary way and your power to sway opinions an control narratives is far greater than the u.s. government's power ever has been. so i notice a trend today. there is a growing desire from many colleagues to make you the arbiters of truth. they know you have this power and they want to direct that power for their own political gain. mr. zuckerberg, since facebook was my first love, i am going to direct questions at you. this isn't a trick. do you believe in the spirit of the first amendment, basic liberal values in. >> yes, absolutely. >> see, my colleagues can't infringe on the first amendment. american people in their speech are protected from government as they should be, my colleagues, this administration, they can't silence people they disagree no matter how much they want to. but i think they want to. i heard democrats complain about
misinformation which they mean political speech they disagree with. they complained that university content is still up. i heard them accuse conservative veterans of being tinfoil hat wearing extremists and opinions on climate change that they disagree with should be taken down. this is quite different from the republican complaint that illegal content needs to be addressed. there is a growing number of people in this country that don't believe in free speech and free debate. i promise you the death of the first amendment will come when the culture no longer believes in it. when that happens it becomes okay to jail or investigate citizens through speech like in canada and throughout europe. their culture turned against free speech. you all sitting here today as witnesses are part of the culture. you can stand up for the spirit of open debate and free speech or you can be the enemy of it. it's clear many want to weaponize your platforms to get you to do their bidding for them. mr. zuckerberg, do you think if your place -- it's your place to be the judge of what is true
when it comes to political opinions? >> congressman, no, i don't believe we should be the ash terb of truth. >> thank you. and, look, i promise you this. as long as you resist these increasing calls from politician toss do their political bidding for them, i will have your back. when you don't, you become an enemy of liberty and long standing american tradition. you might agree in principle with what i said. mr. zuckerberg, you clearly do. i have a feeling the others would answer it as well. the fact remains the community strpds on social media platforms are perceived to be applied unequally and with blatant bias. mr. doors ain one example, i saw a video of -- from project veritas taken down because they confronted a facebook executive on his front lawn. i can show you a video of cnn doing the exact same thing to an old woman who was a trump supporter in her front yard. i have looked at both videos. apples to apples comparison. cnn remains up.
project veritas was taken down. i have a feeling you will tell me you have to look into it. >> i don't have an understanding of the case, but i would imagine if we were to take a video like that down, it would be due do x'ing concern, address. >> the address was blurred out. you don't have the case in front of you. i get this. there are countless examples like this. i just found that one today. there are countless examples like this. so even if we agree in principle on everything i went over, you have lost trust. and you have lost trust because this bias is seeping through. we need more transparency. we need better appeals process and more equitable application of your community guidelines because bev to root out political bias. that /* in these platforms. i think -- and i talked with a lot of you offline or at least your staff and i think there is some agreement there and i hasn't heard anybody ask you what you are doing to achieve
these goals. so i will allow you to do that now. maybe mr. zuckerberg, we'll start with you. >> to achieve which goals? >> more transparency, more feeling that -- better appeals process for content taken down, more equitable application of community guidelines. - >> so for transitio transparency we issue enforcement reports on how -- what prevalence of harmful content of each category from terrorism to incitement of violence, child ex ploy toigs, how much there is and how effective we are at fining that and stats around that. for appeals, the biggest thing that we have done is set up an independent oversight board which is staffed with people who all have a strong commitment to free expression for whom people can appeal to them and that group makes a binding decision, including overturning several of the things that we have taken down and telling us that we have
to put them back up and we respect that. >> gentleman's time is expired. chair now recognizes last but not least my fellow pennsylvanian, mr. joyce. you are recognized for five minutes. >> thank you for yielding. and thank you, mr. chairman, and the ranking members for convening this hearing. i thank you all. it's been a long day. but this is an incredibly important day. we have heard consistently during this hearing about alarming accounts of content policing, censorship, and even permanent deplatforming of individuals. i have also been concerned about the lack of transparency and consistency in facebook's application. facebook's own standards. as you mentioned, i am a representative from pennsylvania, and in my district facebook shut down the personal pages of walt ta hallski and
stharl schaefer as well as a facebook page they administered in historic gettysburg, pennsylvania, and this occurred without warning. since the pages were taken down in december, these pennsylvanians haven't received an acceptable answer from facebook about why they were banned nor have they been given the opportunity to appeal -- >> i am not specific can the details. in general i agree that building out a better appeals process and better and more transparent communication to people about why specific decisions were made is one of the most important things that we need to do next. and that's one of the big things in our roadmap for this year and next year. i hope we can dramatically improve those experiences. >> mr. zuckerberg, may i get from you a commitment that a nor
concise and transparent appeals process will be developed? >> congressman, yes, we are working on more transparent communication to people and more of an appeals process as part of our product now, like i just said. >> and will you commit to getting my constituents answers to why they were banned? >> congressman, i can certainly have my team follow up with them and make sure that we can do that. >> thank you for that. i am also concerned by potential partisan bias in facebook's enforcement of its content policies shutting down the adams county republican committee facebook page strikes me as an infringement on speech and that is normally protected in the public domain. mr. zuckerberg, does facebook maintain data on how many democrat and republican county committee pages you have banned from your platform?
>> no, congressman, we don't. we don't generally keep any data on whether the people who use the platform are democrats or republicans. >> then let me -- i know time is running short here. mr. zuckerberg, you say you have not maintained that data. would you consider gathering such data to verify that there is no political bias in your enforcement algorithms? >> congressman, i am not sure that that's a great idea. i don't know that most people would want us to collect data on whether they are a democrat or a republican and have that be a part of our -- >> i think there is a huge disparity as i represent pennsylvania and i think that that data would be appreciated if shared with us in a fair manner. my next question is to mr. dorsey. does twitter maintain data on the political affiliations of those that you block? >> no.
>> have you determined that any political bias is necessary for your enforcement? >> not sure what you mean, but no. >> i think that these discussions today are so important. i think that you all recognize that the platforms that you represent have developed an incredible pablt for americans to connect and contact. but these free speech that we hold so dear to us must be maintained. again, i thank the chairman, i thank the ranking member for bring us together and allowing us to present what i feel are sincere concerns to you. thank you, mr. chairman, and i yield. >> i thank the gentleman yields. everyone who wanted to ask a question has asked one. i thank you for your patience
today. i request unanimous request to enter this into the record. letter from asian americans advancing justice, a letter from the leadership conference on civil and human rights, a letter from new america's open technology institute, a letter from new york small pharma limited, a statement from the alphabet workers union, letters from national black justice coalition, a letter from seeks for justice, a letter from state ags, a letter from the computer and communications industry association, a letter from avaaz opening statement from anna eschu, neil freed of digital frontiers advocacy, from the music, a letter from the disinfo defense league, a letter from consumer reports, a report from the center for enring hate, letter from transparent internet, let fundraiser the seek american legal defense and education fund, letter from gun
violence survives, faces of tech harm congress, youtube, letter to facebook, letter to twitter, a longitude al analysis of youtube's promotion of conspiracy videos, a let fundraiser the alliance for safe online pharmacies, a ccia statement, a comment by donovan et al from the technology and social change team, wall street article titled facebook shuts down sites to make sites less devicetive, cost americans 4.2 billion, a global research project report and opinion article titled gaga is not cracking down on the most dangerous drug in mark, mit technology review article how facebook got addicted to spreading misinformation, an article from the independent, an article from the new yorker, a letter from the coalition of safer web, a "new york times" article tech companies detect the surge in online videos of
child sex bus, thank you for posting smokers lessons for regulating social media, an article from in premise, an article from the atlantic, a "new york times" article square jack dorsey's pay service is withholding money matterer merchants need, a response letter from twitter to rogers, a response letter from facebook to rex rolkers, an article from in gadget, a letter regarding span irn language misinformation, data from the centers for disease control the national survey on drug hughes and health and the lima stone and wang regarding teen mental health a report from the house committee on veterans affairs. without objection, so ordered. i want to thank our witnesses today for appearing. we appreciate it. we appreciate your patience while you answered those questions from all members. i hope you can take away from this hearing how serious we are
on both sides of the aisle to see many of these issues that trouble americans addressed. thank you for being here today. i want to remind all members that pursuant to committee rules they have ten business days to submit additional questionsf.yk the record to be answered by the witnesses who have appeared, and i would ask that each witness respond promptly to any questions that you may receive. at this time this hearing is adjourned. >> mr. chairman -- next a look at access and distribution of the covid-19 vaccine. this hour-long event was posted by the clinton foundation. weeknights in month featuring american history tv programs as a preview of what's available every weekend on
c-span3. lonnie bunch is the first african american and historian to serve as secretary of the smithsonian institution. tonight his discussion with documentary filmmaker ken burns about the complexities of telling america's story. watch tonight at 8 eastern and enjoy american history tv ever weekend on c-span3. next a look at access and distribution of the covid-19 vaccine. this hour-long event was hosted by the clinton foundation. >> hi, everyone. thank you so much for joining us this afternoon for two conversations i am particularly excited to be able to be a part of and am especially grateful to our participants, all of whom are incredibly busy for taking time out of their busy days to talk with us about the current state of covid-19 vaccinations across ourou