Big Tech and the Online Child Sexual Exploitation Crisis (QFR Responses)

Question"Is your company using safety technology to detect and prevent live video child sexual abuse on your platforms and apps that allow users to stream or share live video?"

ResponseSays any CSEAN violation is one too many and cites heavy investment in technology to detect and remove such content.

QuestionCites a WSJ report that Instagram helped promote accounts trading underage-sex content and asks what personal steps prevent recurrence.

ResponseSays Meta wants safe, age-appropriate experiences and is committed to protecting young people, calling it an ongoing industry-wide challenge.

QuestionCites whistleblower Arturo Bejar's warnings about teen harms and asks how Meta ensures employee safety concerns are taken seriously.

ResponseSays Meta values employee input, encourages internal challenge, and has implemented many employee safety proposals, though not all.

QuestionHow many users under 18 does Meta have on its platforms?

ResponseAbout 6% of US Instagram accounts and 1% of US Facebook accounts self-report as under-18 teens; Meta describes its automated/manual age-checkpoint review process and Yoti age-verification partnership for identifying underage users.

QuestionHow many users under 13 (below Meta's minimum age) does Meta estimate are on its platforms?

ResponseMeta does not provide a current estimate, but cites removing 4.8 million Facebook and 1.7 million Instagram accounts in the last two quarters of 2021 for failing the minimum age requirement.

QuestionHow many users use Family Center parental supervision tools?

ResponseMeta describes Family Center's purpose and cites retention statistics (e.g., 90%+ of parents/teens keep supervision tools 30 days after adoption) rather than providing a raw user count, and pivots to advocating for app-store-level parental approval legislation.

QuestionWhat is Meta's annual budget specifically for addressing online child sexual exploitation?

ResponseMeta says it does not segment its budget by harm type and cannot provide a precise CSE-specific figure, citing only its overall ~$5 billion 2023 safety and security spend.

QuestionHow many Meta employees work specifically on addressing online child sexual exploitation?

ResponseMeta declines to give a CSE-specific headcount, citing only its overall ~40,000-person safety and security workforce.

QuestionHow did Meta decide 13 was the appropriate minimum age?

ResponseMeta says the age-13 threshold is industry-standard and driven by COPPA compliance plus internal/external expert consultation, without citing specific research justifying 13 as the appropriate age.

QuestionWhat US legal obligations require Meta to ensure platform safety for children before launch?

ResponseMeta cites COPPA compliance and an internal child-development advisory board (Yale Center for Emotional Intelligence, ConnectSafely, Sesame Workshop, etc.) rather than identifying a specific pre-launch legal safety-testing obligation, and pivots to advocating for app-store parental approval legislation.

QuestionWhat are the default privacy/content/messaging settings for under-18 users, can teens change them without parental awareness/consent, and how many changed them in 2023?

ResponseMeta details default private-account, restricted-DM, and content-visibility settings for under-16/under-18 users, notes that supervised under-16 teens need parental approval to loosen privacy/DM settings, and cites adoption-retention percentages rather than a literal 2023 count of setting changes.

QuestionWhy does Meta set different defaults for 16-17-year-olds vs. under-16s, and did any employees object to this approach?

ResponseMeta attributes the age-tiered defaults to expert consultation on developmental differences, and does not address whether any personnel objected to the approach.

QuestionGiven reporting that under 10% of teens enabled parental supervision by end of 2022, what data does Meta have on the efficacy of its parental controls, and what outreach does it do to promote them?

ResponseMeta does not provide specific efficacy studies/data as requested, instead describing its outreach efforts (Screen Smart events, Family Center, TTC Labs, ad campaigns) and pivoting at length to its proposed app-store-level parental-approval legislative framework.

QuestionWhat is Meta doing to comply with the EU DSA, UK Online Safety Act, and Australia's Online Safety Act, and will it extend any resulting child-safety improvements to the US?

ResponseMeta describes DSA transparency reporting and a new EU compliance function, cites Australia's child-sex-abuse-material transparency notices and industry codes, says it is still awaiting UK Ofcom guidance, and does not commit to extending any resulting child-safety measures to the US.

QuestionWhy could outside academic researchers detect Instagram's pedophile networks when Meta could not, and does Instagram still connect/promote such networks today?

ResponseMeta describes a task force created after the WSJ report, technology to prevent suspicious adults from being recommended to teens, Group/Page removal figures, and its Lantern industry-sharing program, without directly explaining why outside researchers found the networks first.

QuestionWhat is Meta's average response time to reports of child exploitation?

ResponseMeta does not provide a response-time figure, instead referring back to its Question 8 answer about proactive detection technology.

QuestionHow long is it acceptable for a reported CSAM livestream to continue on Instagram?

ResponseMeta says any violation is "one too many" and details its detection tools (PDQ, TMK-PDQF, PhotoDNA, Google Content Safety API) and enforcement statistics, without answering how long a reported livestream is allowed to continue.

QuestionWhy doesn't Meta suspend accounts pending investigation once reported for distributing CSAM?

ResponseMeta says it promptly disables accounts once violations are confirmed but argues that automatically suspending accounts based on a report alone (before investigation) risks overenforcement and abuse of the reporting system.

QuestionWhat accountability should Meta face for failing to respond to child exploitation reports?

ResponseMeta does not propose any specific accountability mechanism, instead citing its NCMEC reporting volume (26 million reports in 2022) as evidence of industry leadership.

QuestionGiven Meta's own commissioned human-rights assessment warning that encryption facilitates child sexual abuse and exploitation, how many fewer CyberTips does Meta expect to submit to NCMEC once E2EE is implemented?

ResponseMeta does not give a projected reduction figure, instead describing its Human Rights Impact Assessment's 45 mitigation suggestions, AI-based detection in encrypted environments, and a new default-on nudity protection feature for teens.

QuestionDoes Meta acknowledge that default end-to-end encryption will lead to more undetected child exploitation?

ResponseMeta declines to directly answer, referring back to its Question 10(a) response.

QuestionHow will Meta assess whether encryption makes its platforms more appealing to child sexual exploitation offenders?

ResponseMeta again declines to directly answer, referring back to its Question 10(a) response.

QuestionWhy did Meta reject internal employee proposals—like exempting teen accounts from encryption and not recommending minors to adults via "People You May Know"—meant to protect kids from encryption's risks?

ResponseMeta says it "takes input seriously" but "may not adopt every proposal," defends encryption as safety-compatible, and lists mitigating features (restricted teen DM defaults, suspicious-account non-recommendation, nudity protection, parental supervision) without directly explaining why the specific employee suggestions (no encryption for teen accounts, no "People You May Know" minor-adult recommendations) were rejected.

QuestionHow has Meta measured the impact of its ~21,000 layoffs (including trust and safety staff) on its efforts to stop online child sexual exploitation?

ResponseMeta disputes that trust and safety was disproportionately affected by the 2023 layoffs but provides no studies, reports, or data measuring the cuts' actual impact on child exploitation efforts, despite being asked to do so.