Protecting Kids Online: Facebook Instagram and Mental Health Harms

QuestionMarkey asks Davis to commit, yes/no, to supporting his newly reintroduced KIDS Act (banning follower counts/autoplay/push alerts for kids, limiting influencer ads, prohibiting harmful-content amplification); accuses Facebook of "delay and obfuscation" when she won't.

ResponseDavis expresses general openness to "updated Internet regulations" but does not commit to supporting the specific bill.

QuestionCiting the 2014 "emotional contagion" study and recent leaks showing similar effects on teen Instagram users, Thune asks what Facebook will do to counter it.

ResponseDavis distances Facebook's research from "emotional contagion" framing specifically, but describes a planned "nudges" feature to redirect users away from negative content spirals.

QuestionLuján presses, repeatedly, for a straight yes/no on whether Facebook will release the underlying dataset (minus PII) behind its internal teen body-image research.

ResponseDavis never gives a clean yes/no; repeatedly cites privacy obligations and says Facebook is "looking to release more" without committing.

QuestionLuján asks, yes/no, whether Facebook tests algorithm changes' impact before rollout, and specifically whether it tests changes for their effect on user growth/revenue.

ResponseDavis says this isn't her area of expertise both times; doesn't confirm or deny growth/revenue-impact testing.

QuestionLuján asks, yes/no, whether Facebook tests platform changes for their effect on users' propensity toward violent/hateful language, and separately toward self-harm ideation.

ResponseDavis doesn't answer the hateful-language question directly; on self-harm, cites the ~0.5% figure of young users tracing first suicidal thoughts to the platform and describes the suicide-prevention reporting flow.

QuestionBlumenthal presses on who specifically decides whether Facebook's internal research gets made public, directly asking if it's Zuckerberg.

ResponseDavis doesn't name Zuckerberg or anyone else; says it's a collective decision and points to existing transparency measures (transparency reports, oversight board, human rights impact assessments) as evidence of good faith.

QuestionCiting Molly Russell's death and an internal April 2021 document quoting a Facebook engineer saying "no one has decided to dial into eating disorders," Blumenthal says Facebook knowingly let its recommendation system push users toward eating-disorder/self-harm content.

ResponseDavis lists existing mitigations (suicide-prevention partnerships since 2006, eating-disorder reporting flow, resource pop-ups) without directly addressing the "no one has decided to dial into eating disorders" internal quote.

QuestionBlumenthal presses Davis to confirm she's not committing to concrete steps or a timeline on eating disorders/suicidal ideation, and dismisses her described mitigations as "baby steps."

ResponseDavis acknowledges the problem exists "within our society" and on the platform, declines to give a timeline, and describes two in-development features: "nudges" toward uplifting content and a "take a break" prompt for heavy/rabbit-holing use.

QuestionBlumenthal asks how long the Instagram-for-kids pause will last and who — specifically whether Zuckerberg — will decide.

ResponseNo timeline given; Davis again declines to name a decision-maker, saying it's a collaborative process.

QuestionBlackburn asks what Facebook does with data from its "digital experience surveys" of kids, how long it's retained, and demands Facebook submit its parental-consent form for the record.

ResponseDavis says the most stringent privacy protections apply and parental consent is obtained, but doesn't answer the data-retention question and only agrees to "take the request back" on producing the consent form (no commitment to actually produce it).

QuestionLummis contrasts Zuckerberg's earlier congressional testimony (social apps have "positive mental health benefits") with internal research Facebook allegedly withheld, and asks directly whether Facebook is withholding negative-mental-health-impact studies.

ResponseDavis pushes back that the WSJ reporting was "one sided," cites comparable outside research (Harvard, Pew, Berkeley), and gives the "8 out of 10 report a neutral/positive experience" figure; doesn't directly answer the withholding question.

QuestionLummis asks whether Facebook uses data on the "2 out of 10" users with negative experiences to adapt its product for kids being harmed.

ResponseDavis describes the same "nudges" toward uplifting content approach mentioned elsewhere in the hearing.

QuestionCruz asks whether Zuckerberg was personally aware of the teen-harm research per WSJ reporting, and whether Davis herself has ever discussed it with him.

ResponseDavis doesn't confirm Zuckerberg's specific awareness; says she personally doesn't recall discussing this particular research with him.

QuestionCruz repeatedly presses Davis on the WSJ-reported "13%/6% trace their desire to kill themselves to Instagram" finding — is it accurate, has the underlying research been fully released, and how can "not bombshell" be squared with that statistic.

ResponseDavis repeatedly disputes the WSJ's "13%/6%" framing as a "misunderstanding"/"mischaracterization," citing a 0.5% figure instead; never gives a yes/no on whether the full underlying research has been released.

QuestionCruz asks (1) whether Facebook changed any policies in response to the suicide-ideation research, and (2) whether Facebook has ever quantified how many children have died by suicide attributable to its products, pressing repeatedly for a yes/no.

ResponseDavis never confirms a specific policy change tied to this research; on quantifying child deaths, she repeatedly says the research isn't "causal" and that such quantification would require long-term study Facebook hasn't done — effectively confirming, under Cruz's characterization, that no such count exists.

QuestionLee asks, yes/no, whether Facebook/Instagram allow businesses to target adult-themed ads at 13-17-year-olds.

ResponseNo clean yes/no; Davis describes the three allowed ad-targeting parameters (age, gender, location) and some prohibited categories (e.g., weight-loss ads).

QuestionLee narrows to sexually suggestive/explicit ad targeting specifically, then asks what Facebook's process is for determining which ads are age-appropriate for kids.

ResponseDavis doesn't directly confirm or deny sexually-explicit ad targeting; describes categorical ad prohibitions and offers to provide the "full list" afterward rather than in the hearing.

QuestionLee asks whether Facebook/Instagram's App Store age ratings (12+ on Apple, "Teen" on Google Play) came from Facebook itself, why there's a disparity between those ratings and the actual (sometimes exploitative/pornographic-adjacent) content available, and — pointedly — whether Davis personally believes those ratings are appropriate.

ResponseDavis says Facebook doesn't set its own App Store age rating (submits a form, the platforms assign it), doesn't say whether she thinks the ratings are appropriate, and repeatedly redirects to Apple/Google as the ones to ask.

QuestionBlumenthal directly asks Davis to commit to ending "Finstas" (secret teen second accounts).

ResponseDavis again frames Finstas as teen-driven behavior rather than a Facebook product/feature, and doesn't commit to ending anything. Blumenthal explicitly says on the record this isn't a real answer.

QuestionSullivan asks directly whether Davis was aware of the teen-suicide-statistics research, then challenges her framing that Instagram use "helps" more than harms, calling it "quite remarkable" and pressing for actual evidence.

ResponseDavis confirms awareness of teen mental health issues, repeats the "11 of 12 issues, more teens found it helpful than harmful" framing, stresses the research is not causal, and again describes the "take a break" and "nudging" features.

QuestionSullivan cites China's regulatory time-limits on teen social media/gaming use and asks whether the U.S. government should consider similar mandated limits, since he doesn't trust the industry to self-regulate.

ResponseDavis argues parental controls (not government mandates) are the right approach, and cites COVID-era connectivity benefits as a complicating factor.

QuestionBlumenthal offers Davis a closing opportunity to add anything before the hearing adjourns.

ResponseDavis uses her closing remarks to point at TikTok and YouTube as also serving under-13 users, framing the under-13 access problem as an industry-wide issue rather than Facebook-specific.

QuestionKlobuchar follows up in writing on who decides whether/when to unpause Instagram Kids, what criteria will be used, who Facebook is consulting, and whether the Board has a role.

ResponseFacebook names Adam Mosseri (Head of Instagram) as the one who will ultimately decide whether to resume Instagram Kids, in consultation with Meta leadership — directly answering, in writing, the "who decides" question Davis repeatedly declined to answer live.

QuestionKlobuchar asks for a per-platform breakdown of ad revenue per user, and specifically how much comes from users under 18.

ResponseFacebook declines to provide a platform-level or under-18 revenue breakdown, citing only its already-public aggregate ARPU/ARPP figures.