Disinformation Nation: Social Media’s Role in Promoting Extremism and Misinformation

QuestionLatta recounts the Amanda Todd case — a 15-year-old who died by suicide after being harassed with illegal images of herself shared on Facebook — and repeatedly asks, as a yes/no question, whether Facebook bears any responsibility for her death.

ResponseZuckerberg never gives a yes/no answer on Facebook's responsibility for Amanda Todd's death specifically, instead speaking generally about the company's AI/content-moderation investment in child exploitation content.

QuestionLatta asks what specific Section 230 reforms Zuckerberg supports, given his testimony citing child exploitation as a category platforms should be held accountable for, and presses on transparency-reporting frequency.

ResponseZuckerberg proposes two Section 230 changes for large platforms: mandatory transparency reports on harmful-content prevalence (naming child exploitation and terrorism as example categories), and conditioning immunity on having effective systems to moderate clearly illegal content like sex trafficking and child exploitation.

QuestionAsks how a proposed free under-13 service would monetize users, whether it targets children, and if he'd let his own kids use it.

ResponseSays Meta is still early in exploring the under-13 service concept and currently doesn't allow under-13 use.

QuestionChallenges the consistency between discouraging passive screen time and designing to increase it, and asks if the goal is maximizing time on-platform.

ResponseSays algorithms target meaningful interactions, not maximizing time spent, and that teams aren't given time-spent goals.

QuestionUpton asks what resources Meta devotes to combating illegal drug sales on its platforms, given the practice remains a problem despite violating policy.

ResponseZuckerberg cites over 1,000 engineers on AI "integrity systems" and 35,000+ content reviewers, and frames illegal-drug-sale content as a case where large platforms should be expected to build effective moderation systems.

QuestionCiting initial reports that the Boulder, Colorado mass-shooting suspect had been bullied, including online, Upton asks what process exists for parents/families to pursue anti-bullying efforts on Meta's platforms, and what changes to Section 230 Zuckerberg would suggest.

ResponseZuckerberg says bullying content is often not clearly illegal, making it a key example for why Section 230 reform is needed beyond just illegal content, and frames "stop[ping] people from bullying children" as one of Meta's primary use cases for that broader moderation authority; reiterates his three-part Section 230 reform proposal.

QuestionEshoo asks, as a yes/no question, whether Zuckerberg will commit to redesigning Meta's products to eliminate their focus on "addicting users to your platforms at all costs" for ad-revenue purposes.

ResponseZuckerberg does not give a yes or no answer and is repeatedly cut off before completing a substantive response; Eshoo concludes his non-answer amounts to "no."

QuestionReferring to anti-Asian-hate-associated hashtags (e.g. COVID-origin slurs) circulating on social platforms, Matsui asks why Facebook still allows them and presses Zuckerberg specifically for comment after Dorsey answers.

ResponseZuckerberg calls anti-Asian hate a priority issue and says Facebook removes hateful hashtag content that violates its hate speech policy.

QuestionCites studies showing algorithms promote divisive, hateful content because it boosts engagement, and asks what's being done to fix it.

ResponseDisputes that the goal is keeping people on the service, saying the aim is meaningful social interactions.

QuestionCastor cites the documentary "The Social Dilemma," a 2019 JAMA pediatric study linking adolescent depression risk to social media hours, and 2019 HHS research on rising suicide rates among 10-to-14-year-olds, asking whether Zuckerberg is aware of each, and whether he's aware of research linking self-harm hospitalizations to addictive app design.

ResponseZuckerberg says he hasn't seen "The Social Dilemma" but is "obviously familiar" with the JAMA study's general subject, says he's not aware of the specific HHS suicide-rate research, and gives no response at all to the final self-harm-hospitalizations question — the official transcript marks it "[No response.]" — before Castor moves on to Pichai.

Question"Mr. Zuckerberg, how much advertising revenue does Facebook— do you make from behavioral surveillance advertising targeted towards kids under age 13?”

Response"Congresswoman, it should be none of it. We don’t allow children under the age of 13—On the services that run advertising."

Question"[A]re you saying that there are no kids on Instagram under the age of 13 right now?"

Response"Congresswoman, children under the age of 13 are not allowed on Instagram. When we find out that they are there we remove them."

QuestionAccuses him of profiting by hooking users to the platform and asks if he agrees the product is designed to create addiction.

Response"Congressman, no. I don’t agree with that.”

QuestionAsks whether kids commonly lie about their age to bypass the 13+ rule, and whether under-13s can currently use Instagram.

ResponseAcknowledges kids lie about age, describes detection systems and an explored under-13 service with parental controls, and confirms current policy still bars under-13 use.

QuestionClarke asks why Facebook continues to allow racially discriminatory ad targeting despite past updates to its "special ad audience" tools.

ResponseZuckerberg cites steps taken to eliminate racial-affinity ad targeting and ongoing consultation with civil rights experts.

QuestionCiting the El Paso and Atlanta mass shootings as examples of hate-fueled violence linked to online content, Cárdenas asks whether Zuckerberg believes Meta has done enough to combat the spread of hatred and incitement to violence on its platforms.

ResponseZuckerberg says Meta has done more than any other company but acknowledges the problem persists and more needs to be done.

Question"Do you believe that your platform harms children?"

ResponseSays he doesn't believe the platform harms children, citing design goals around well-being and connection and the under-13 ban.

QuestionAsks about the 2.5 million suicide/self-injury posts removed in Q4 2020 and what policies expanded reviewer capacity over time.

ResponseSays AI now identifies roughly 98-99% of hate speech, terrorist content, and suicide-related content Meta removes.

QuestionCraig describes a specific incident where Facebook took down a colleague's trans-flag video as hate speech while leaving up another member's transphobic video, asks Zuckerberg to confirm the facts and whether the disparate enforcement seems right, and asks whether he'd support a mandatory Office of Civil Rights reporting to the CEO.

ResponseZuckerberg says he's unaware of the specific trans-flag/transphobic-video incident, agrees the described disparate enforcement doesn't "seem right" in the abstract but won't confirm the specifics, acknowledges content-moderation mistakes happen and Meta won't ever be "perfect," and cites hiring a VP of civil rights rather than committing to a CEO-reporting Office of Civil Rights.

QuestionTrahan describes manipulative, attention-capturing design features in kids' apps like Messenger Kids and the reported Instagram for Kids, and asks Zuckerberg directly whether the planned Instagram for Kids will include endless scroll and beautifying filters, plus how he'll personally handle his own children craving screen time over face-to-face interaction.

ResponseZuckerberg says Instagram for Kids' design (including whether it will have endless scroll or beautifying filters) is not yet finalized, and points to Messenger Kids' parent-controlled model as the template being considered.

QuestionFletcher asks whether a federal legal standard for defining domestic terror/hate organizations, similar to the State Department's foreign-terrorist-organization designation, would help Meta better track and remove harmful domestic content.

ResponseZuckerberg says he's not sure a federal standard is needed, noting Meta already classifies domestic white-supremacist organizations, militias, and QAnon-linked networks at the same severity level as foreign terrorist organizations.

QuestionMcKinley confronts Zuckerberg with fresh staff-identified examples of oxycodone, Ritalin, Xanax, and Adderall being sold on Instagram — three years after Zuckerberg apologized for the same problem on Facebook — and repeatedly asks whether Zuckerberg should be held personally liable, comparing it to liability for retailers selling cigarettes to minors and manufacturers held liable for the opioid crisis.

ResponseZuckerberg does not answer the personal-liability question directly, instead arguing that at Meta's scale it's inevitable some illegal-drug-sale content will be missed (comparing it to a police force that can't stop every crime), and pushes back on the characterization that Meta is "studying the problem" rather than acting.

QuestionCiting her DETOUR Act targeting "dark patterns" — deceptive interface tactics used to extract personal data — Blunt Rochester asks whether Zuckerberg would oppose legislation banning such manipulative design, and whether he'll commit to including deceptive design choices in Meta's own internet-regulation proposals.

ResponseZuckerberg says the anti-dark-patterns principle "makes sense" but declines to commit, saying he'd need to think about it and that other regulatory areas feel more urgent to him.

QuestionCiting NCMEC CyberTipline data showing Facebook responsible for 20.3 million child-exploitation reports (far more than Google, Twitter, or MindGeek), Griffith asks, yes or no, whether Facebook has a child exploitation problem on its platform.

ResponseZuckerberg attributes Facebook's high NCMEC report volume to proactive over-flagging rather than a uniquely high rate of child exploitation content, arguing high report numbers should be seen as evidence of diligence rather than criticized.