2014–2026
539 passages quoted from Meta's internal records, court filings and congressional testimony, in date order. Each is labelled with its source.
"Over the course of 2024, our work will ladder to the following four outcomes… Drive acquisition of teens" "For both existing and new products, we will evaluate their potential for teen resonance as well as the ways that they unlock user value." "Acquiring new teen users is mission critical to the success of Instagram."
"Prioritization. I said this in my last round of feedback, but I think it's worth repeating: Adam has helped the lnstagram organization focus and it's leading to better results. Adam has been extremely clear about our focus on teens, and helped us maintain a shorter, more consistent list of top priorities. Adam has consistently pushed for hard trade-offs to achieve our goals and been willing to accept the consequences, e.g. pushing for more original content even at some cost to Engagement." "lnstagram's messaging product will have a major impact on teen usage, and therefore our chances of winning with teens, over the next 5 years. We need to close gaps on messaging data usage for ranking, group messaging, and PREQ to compete effectively, but we haven't made material progress on any of these issues in 2023. Adam is uniquely positioned to help push for better execution because there are three top-level product groups are involved (Instagram, Messenger, and Family Foundation)."
Child sexual exploitationFriend/Account Suggestions
"Prior to the Task Force, we had some limitations the recommendations between minors and potentially suspicious adults, but they didn't cover everything and we're always bi-directional. As part of the task force, we rolled out bi-directional protections to, most notably, PYMK recommendations. But they didn't exist there prior to the Task Force." "My understanding is that, prior to the CSTF, we were recommending minors to potentially suspicious adults and vice versa in PYMK. Closing that gap, which we did, was accomplished as part of the CSTF."
"Even if things are going well in terms of topline trends, Adam makes sure we don't take our eye off the ball of where the puck is going (e.g., lots of urgency around DM teens concerning trends)."
"My specific goals were to (1) stabilize teen growth… Teen growth in developed markets, though not yet healthy, is looking up and has stabilized in roughly half of those countries. MAU continues to fall, but we we have a credible path to stabilize for the first time."
"Adam continued to steer IG to focus on the key priorities, especially with IG Teens Growth and Threads. Such consistency and repeated drum beating helped make things easier for all of us to collaborate and prioritize."
Child sexual exploitationFriend/Account Suggestions
Gabriel ████: "(l) Yes on PYMK, but with the specifics of a/ MaCSA>0.3 to Minor: Filtered b/ Minor to MaCSA >0.3: Filtered except where 2 year age gap c/ Minor to MaCSA >0.6: Filtered with no age gap (2) There were some filters in place prior to taskforce using UnIS criteria and MaCSA previously, but these were still leaving a large number of exposures in PYMK between MaCSA<>Minor so the Taskforce work was plugging the gaps" Sadaf ████: "[J]ust want to make sure i understand 'Minors are not recommended to potentially suspicious adults' was *not* in practice before the task force" Kaitlin ████: "It sounds like there was some filtering of not recommending minors to suspicious adults AND vice versa before the task force, but it was not comprehensive and the task force made it better. What's definitely wholly new as of the task force is not recommending suspicious adults to each other (or anyone)... we were already doing some of Bullet #2 prior to the task force, just poorly… 'There were some filters in place prior to taskforce using UnIS criteria and MaCSA previously, but these were still leaving a large number of exposures in PYMK between MaCSA<>Minor so the Taskforce work was plugging the gaps')... My understanding… is that we had no filtering of recommending potentially suspicious adults to each other prior to the TF. We DID have some (though inadequate) filtering of sus adults to minors and vice versa…. So there are four types of recommendations: 1) Minors to Sus Adults 2) Sus Adults to Minors 3) Sus Adults to Sus Adults 4) Sus Adults to ANYONE My understanding is both 1 & 2 existed before the TF but needed improvement and that 3 & 4 were net new." Davis: "prior to task force it was limited to minors, post task force it was to anyone – that is correct" Gabriel ████: "Yes. there was bi-directional risky adult <> minor filtering prior to the taskforce, though the interventions in each direction did not use the same filtering criteria. Adults filtered from Minor PYMK recs using MaCSA, migrated from previous IIC classifier. In the taskforce audit we found that this intervention was not filtering all recommendations and needed to re-implement it to ensure it was not leaking exposures. - Minor -> Adult using UnIs (legacy definition of risky pair). In the taskforce, we migrated this filter to the same MaCSA threshold as used elsewhere, which added a much broader volume of exposures and required escalation with Growth. The outcome of this escalation was that we created a 2 year age gap carveout, but implemented the MaCSA filtering for all other minor-> risky adult PYMK recommendations. NET NEW delivered in taskforce - "MaCSA 0.6 -> Filter from all users in PYMK" - MaCSA 0.3 <> MaCSA 0.3 Filtering in PYMK"
PLAINTIFFS’ CHARACTERIZATION
Child sexual exploitationFriend/Account SuggestionsContent Algorithm
Addiction & problematic useContent Algorithm
████: "They should audit every surface team's goal across the company, think about whether we would be comfortable telling the world that we are optimizing for x for their kid, and when the answer is no, come up with a defensible alternative." ████: "imagine going out in public and telling people we rank their kid's facebook by what makes them consume the most content IDs lol" ████: "this will be a public conversation and we are not ready for it" ████: "and then when they hear what things keep their kid consuming the most content IDs … we're off to the races on a really fun conversation."
"In 2024, the organic IG Reels watch time budget will increase from -6.4% to -7.8% to account for the impact of Reels ad load increases on Reels consumption. The impact of Reels ads on engagement is typically immediate on Reels organic watch time, but takes time to materialize into Sessions and DAP, which is why the Reels organic watch time was flat in Q4, while the impact on Sessions and DAP continued to grow. In 2024, we will increase the IG Reels organic watch time budget to 7.8% to account for the planned investments in Reels ad load scaling. The corresponding session impact is estimated at -0.08% to -0.22% incremental IG session loss in 2024 after factoring long-term impact, which will still be within the 2.54% upper bound forecast for Instagram sessions. Note: Reels ads drives higher Session and DAP loss on Instagram than on Facebook, and the team's hypothesis is that this is because Reels represent a larger portion of the app's engagement: as of the end of Q4'23, Reels represented 48% of Instagram's time spent, and 20% of Facebook's time spent." "The 2024 Reels ad load scaling strategy continues to focus on ensuring long-term efficiency, with an engagement impact framework efficiency of 92% EIF (significantly higher than the 85% EIF efficiency bar), which will allow us to generate +12% Reels revenue/+0.3% iRev while keeping a high bar on engagement. In Q4 we launched LTV-based ad load which reduced ad load for teens to maximize long-term revenue growth. In 2024 we will continue this focus on thoughtfully increasing ad load to maximize long-term revenue growth. We expect to beat original LRP Reels revenue forecasts even though we expect a lower RME goal (26% vs. prior 31% goal) due to the continuation of the non-regrettable items we've seen in H2'23. The Q1 ad load increase will be focused on non-teens which will go from 15.4% ad load at the end of Q4'23, to 15.8% ad load in Q1'24, and 18% by the end of 2024, while teens will remain constant around 9% ad load."
Suicide & self-injuryDrugsBody image & eating disorders
"A key goal in H1 2024 is to improve Safety for Teens. The top opportunity to do this is to improve enforcements on non-recommendable queries such as 'porn', 'meth', 'self harm', and 'skinny body', a gap that many competitors have already addressed." "Currently, IG shows account results when someone searches for a non-rec query. Some of these accounts have non-rec content, violating content, or sometimes neither, but these accounts can still cause perception impact for the users (teens, parents, adults) - especially when the user has opted to be in a Safe setting (SCC Less)." "Teens use Search to find problematic content, as measured by the fact that Search is a top pathway contributing to teen SWVP and non-rec prevalence. However, parents expect IG to limit the discoverability of age-inappropriate content for teens on search." "Despite SCC Less being our safest setting for teens and IG completely blocking inherently violating queries on search, teens still discover violating and non-rec content/entities through inherently non-rec queries. Results for SCC Less are not significantly safer than SCC Standard results." "It is quite easy to find violating and non-rec content by searching non-rec queries or violating adjacent search queries" "Account-level recall is low, and it is extremely easy for any teen to find pro-eating disorder accounts with very triggering content" "account and content-level recall gaps pose a huge risk to teen safety"
Content Algorithm
"This is so crazy to me…We found groups of new teen users who see Reels just Filled with sexualized media after sexualized…like their whole tab" "The way we found this was so crazy. We looked at Reels VM scores for some users of interest and found that there was nearly a perfect linear correlation between the score and the sexual classifier, to the point that when folks saw the chart they were like 'why is the sexual classifier an input into the Reels' lol… so it seems like an issue with diversity of sourcing, that we source exclusively similar content for these users"
Addiction & problematic use
"Teens' average scores across scales assessing sense of control (attentiveness and capability) indicate they report lower control, on average, over various aspects of their use of IG and TikTok compared to other apps." "Teen average scores on subscales assessing their sense of control over use varied by app and were generally lower for IG and TikTok" "Teen average attentiveness scale scores varied significantly by app; average scores for IG and TikTok indicate teens perceive they, on average, frequently 'spend more time' than desired, 'check [these apps] out of habit', and 'get caught up viewing content' on these apps. Teen average capability scale scores varied significantly by app (though showed a tighter range than attentiveness scale scores by app); average scores for IG, Snapchat, and TikTok indicate teens perceive they, on average, feel somewhat to very confident in their ability to 'control the amount of time they spend', 'reduce how often they check', and 'stop interacting with others' if desired on these apps. 22% of teens in this sample reported that their use of IG 'often' or 'all the time' takes priority over other areas of their life (e.g., work, school, relationships)."
Body image & eating disorders
"US teens in this sample who saw a greater % of appearance-related FIT subtopics reported more appearance comparison (AC)." "Prior research identified a list of FIT subtopics, that when viewed in greater proportions, were associated with worse survey measures of appearance comparison. We manually reviewed content and narrowed the list in H2 '23. Topics include 'Lip Makeup (product)' and 'Fashion Photography.'" "Viewing a greater proportion of these FIT subtopics was associated with more frequent and more negative appearance comparisons on IG, even after controlling for teens' general body satisfaction, age, and gender identity." These findings extend prior AC work by:... Indicating that there is an independent relationship between teens' FIT topic exposure on IG and teen AC that is not likely explainable by teens' other relevant characteristics. This suggests that changing FIT topic exposure may be a way to reduce AC."
Suicide & self-injuryChild sexual exploitationFinancial sextortion
"Financial sextortion cases predominantly impact boys age 14-17" "1 in 8 Of those who reported impact described thoughts of suicide or self harm." "IG is the leading platform reported for initial meeting [45%]" "IG was the leading platform for both threats and reports of distribution"
Addiction & problematic use
"Overall Takeaways/Recommendations: We are currently planning to sunset Take A Break as part of time management consolidation. This aligns with the fact that teens mostly have short sessions. Take A Break is likely not the appropriate market fit time management product" "On average 34% of teen DAU spend 60+ minutes on the app every day." "While we've seen increased adoption of some products we've also seen an increase in the number and percentage of teens with late night time spent at the beginning of the year. The rise did start to taper off but didn't return to the levels that it was at in late 2023." "Teen activity peaks between 6pm and 9pm, with over 40% of teen DAU having sessions that start during that time period. With Quiet Mode starting at 11pm by default, even with the plans to default teens into quiet mode it won't be disrupting their peak hours on the app."
"Goal Drive consideration (& ultimately preference) for IG among teens in service of long term teen MAU growth" "Target Audience Teens(13-17), US)"
Suicide & self-injury
"Threshold value for autodeletion: 0.94."
Child sexual exploitationFriend/Account Suggestions
"FYI I don't think we can stop recommending teens to adults through AYSF on Instagram. If I'm reading the data correctly that'll stop roughly one third of teen follows that come from AYSF, which suggests that we'd be breaking something meaningful. I'm not sure what's going on yet, it may because a lot of teens lie about their age, or did at some point, that they get caught up in this. I was as excited about this idea as anyone, but unfortunately I don't see how we can make it work, so my guidance to the team is to explore other ideas(which unfortunately will mean harder to communicate)."
"People Misrep team, owning the Underage E2E reporting and enforcement flow, focuses on impersonation problems, where age plays a relatively minor factor, resulting in an overall lower priority and understaffed on the Underage domain." "The Underage reporting and enforcement pipeline had been put to Maintenance Mode the past 1.5yr, with the team only conducting bug/Sev fixes and monitoring. Since then, the majority of eng and XFN with context have departed." "The current task force contains: <0.25 Eng, <0.25 DE, 0 PM and other XFN" "IG automated policy to ignore above age users has 70-75% precision (comparing to 95%+ goal), resulting in ~30% reports incorrectly ignored." "Goal was not clearly documented until DS manual audit and investigate. No monitoring or alerting were set-up."
Child sexual exploitation
"17X, by any measure across the industry, is a very, very high strike threshold. That means you could incur 16 violations for prostitution and sexual solicitation, and upon the 17th violation, your account would be suspended."
Bullying & harassmentChild sexual exploitation
"I quickly became aware that it was not uncommon for children in VR to experience bullying, sexual assault, to be solicited for nude photographs and sexual acts by pedophiles." "I was given a legal counterpart to scrutinize everything that I did, to tell me what research I could and couldn't do, and to ensure my research reports would not create 'risk' for Meta should they be publicly disclosed. I was told not to investigate the kinds of harm children were experiencing in VR."
Child sexual exploitationSuicide & self-injury
"On 11 separate occasions between February and December 2025, NCMEC emailed Meta regarding reports in which a child victim expressed suicidal ideation in a chat log, and Meta failed to escalate the report. NCMEC also raised this issue in at least three bi-weekly calibrations with the Meta team... Despite NCMEC's numerous communications to Meta on these issues, Meta's failure to consistently escalate these urgent reports continues to be a problem." "On March 28, 2025, Facebook and Instagram implemented an 'adult classifier' that was described as using account signals to detect users who might be lying about their age... After the classifier became active, NCMEC contacted Meta by email on two occasions after consistently receiving feedback from law enforcement indicating that Meta's 'adult classifier' was generating false positives in which a reported child victim was an adult, negatively impacting law enforcement response and report actionability... but Meta continues to report adults as child victims as a result of inaccurate flagging by its 'adult classifier'." "Despite NCMEC's numerous discussions with Meta regarding persistent quality issues and failure to include substantive information in online enticement and child sex trafficking reports submitted to the CyberTipline, Meta has not resolved these issues. Chat log information included in Meta's reports continue to be routinely and arbitrarily limited, fail to provide sufficient context, or appear to have no nexus to child sexual exploitation... a substantial number of Meta's reports consistently appear to have no nexus to online enticement and/or do not contain sufficient context for law enforcement to initiate an investigation."
Body image & eating disorders
"Appearance comparison is experienced by many in our communities. An estimated 33% of Instagram DAU across 10 countries (FCI + Mexico) compare their appearances to others' on Instagram 'often' or 'always'; 48% 'occasionally' or 'never' compare their appearances. An estimated 48% of teen girls compare their appearances to others 'often' or 'always', and 30% 'occasionally' or 'never' compare their appearances." "Academic research suggests a causal link between Instagram use and appearance comparison and/or body image issues. Longitudinal studies indicate a bidirectional link between social media use and body image issues... Appearance comparison is typically believed to be the link between social media use and body image, particularly on image-based social media platforms." "This is one of the most significant external criticisms we're facing with regard to teen well-being. Regulators and press are pushing us to address social comparison and negative appearance comparison specifically... We've also made our commitment to addressing social comparison a primary proof point in our recent external narratives."
"Sensitive Topic Risks. Asking users on platform about sensitive topics (like appearance comparison) come with a number of risks that must be balanced against likely benefits. These include (1) negative user experiences (i.e., users do not come to our platforms to be asked about high sensitivity topics like AC); (2) concerns about having data on such experiences retained in our systems; and (3) comms risks (e.g., screenshot risks). (4) Unclear or unknown consequences of optimizing for this type of metric (e.g., it may reduce important and healthy content too, like motivational and inspirational posts)." "Benefits may not outweigh these risks: The CWB team has concerns that the benefits of AC survey measurement do not outweigh these risks." "(Mis-) Alignment with Core Company Wellbeing Strategy. Unless anchored in Problematic Use, appearance comparison is not an active core topic prioritized in the x-Meta strategy... it does mean that expected benefits to the company are lower in the overall risk calculation." "Our discussions with policy partners around doing work of this kind strongly advise AGAINST doing work on such sensitive topics with minors UNLESS parental consent is obtained. The message has been: this type of work can be very valuable, but only if done well/ethically, and parental consent is a cornerstone of research with minors, especially if we're asking about health/well-being" "pNAC is not proposed as a metric to be used outside Instagram or to be shared externally." "In November, the NAC survey at the core of the phased validation process was blocked by central survey reviewers for needing additional review given Facebook Files."
Sources include internal documents, public filings and press reports. “Alleged by plaintiffs” denotes the plaintiffs’ characterization, not Meta’s own words. Source labels identify provenance, not a finding of fact.
Meta & Children's Online Safety