Meta's internal documents produced during litigation.
488 records · How documents are sourced
49 records · Meta's internal records
Messages & email · 3 pp
"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 …
1excerpt
Messages & email · 8 pp
████: "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." [Redacted…
1excerpt
Research & report · 5 pp
"Entire Feed Ranking has a great half. Connected Ranking side, we exceeded our global cap-15 session goal and VPV goal. Blending/IFR joint-holdout also exceeds both global and teen cap-15 session goals. This is nearly mission impossible, when we set up goals i…
1excerpt
Messages & email · 1 pp
Praveen ████: "shared: Instagram_Instagram_s_Algorithm_Delivers_Toxic_Video_Mix_to_Adults_Who_Follow_Children_-_WSJ.pdf" Tomas Kabbabe: "It is true that the article seems like an intentional hit piece… But at the same time, some of the content shouldn't…
1excerpt
Messages & email · 2 pp
"[S]o, relevant to the lawsuits, and some foreshadowing from Mark in Q&A just now –– basically the company position is that time spent and engagement metrics proxy user value and we are 'not' just tuning the algorithm to keep people scrolling as much and as lo…
1excerpt
Messages & email · 11 pp
████: "Are the compliance people reading the same litigation and legislation that I am?" ████: "I am asking the same question" ████: does ████ realize how the core algo is set up? Like it's literally just mathematically optimizing for…
1excerpt
Messages & email · 3 pp
████: "I feel like i'm reading totally different litigation and legislation than our org leaders… We literally are optimizing teen feeds for engagement. We can change that, but none of the YIPs actually address this core issue." ████: "[D]o we do …
1excerpt
Messages & email · 2 pp
"Teens are still developing, novelty-seeking and impulsive . . . Cool, so we probably shouldn't have ranking primarily driven by p(eyeballtime) and p(engagement), which is how it currently works."
1excerpt
Research & report · 2 pp
"We want to share a summary of learnings and suggestions for next half in this post... Given sampled teen and overall data, we verified that (1) more teen data leads to consistent teen NE wins and wins for overall NE in most cases; (2) teen data is more effici…
1excerpt
Messages & email · 2 pp
Melissa ████: "I'm not surprised we have bad hashtags. our search safety team is a few employees who look at a list of popular hashtags every week to decide which to filter out :( it's super low tech… We also do very little to make sure handles are appro…
1excerpt
Messages & email · 2 pp
"We wanted to share some (rather concerning) early data insights and next steps, as we looked into recommendability and media quality of US Teens' Top recommended Reels." "Anecdotal evidence indicates that Teens see higher levels of non-rec content. An ex-Met…
1excerpt
Research & report · 19 pp
"The Internal Audit (IA) team has completed the IG Youth Well-being Assurance Audit - Phase 1. This report is rated as 'Significant Improvements Needed' with 3 High, and 14 Medium risk observations." "Key observation themes are summarized below: - Inconsiste…
2excerpts
Messages & email · 2 pp
████: "Can I please see the Mercury project results? ████ led me to ████ who is... not being responsive." ████: "I actually DON'T have access to those results myself, as they were kept pretty confidential... all that is to say, happy t…
1excerpt
Messages & email · 2 pp
Santhosh Vinod: "We're looking for surface integrity representation for AYMF to reduce preference amplification." Dennis ████: "what is preference amplification?" Santhosh Vinod: "Recommender systems can 'rabbit-hole' users into a certain type of conte…
1excerpt
Research & report · 24 pp
"Overall, while IG Reels has generally implemented the SOX controls required for ad products, several areas in content integrity and product features need significant improvements. IA identified 4 High risk, 20 Medium risk, and 3 Low risk observations during t…
1excerpt
Presentation · 4 pp
"Well-being work is similar to privacy or major social issues like Health and COVID19; it needs to be addressed consistently across our platforms and in an end-to-end manner (encompassing both policy/enforcement and product changes). We're not doing this, whic…
1excerpt
Messages & email · 2 pp
Alison Lee: "to be honest, i was dismayed that it felt like Adam's response was that we were 'doing enough', going as far as to question 'how much is enough to invest into integrity?'" Alison Lee: "we've repeatedly had to put p1 improvements like precision tu…
1excerpt
Presentation · 50 pp
"Teens experience harm at higher levels than non-teens" "1. Teens experience a small percent (0.42%) of impressions that are violating current policies but this is ~1.8x higher than for adults and 2-4x higher when weighted by severity across certain violation…
1excerpt
Messages & email · 3 pp
████: "IG and CI teams recently collaborated to outline the policy violating harm IG teens are exposed to. These results are summarized in the doc... We appreciate your feedback and inputs on this and would like to publish this as a note in closed IG/CI …
1excerpt
Messages & email · 2 pp
Meta employees expressed concern with the "well-being challenge" of "content on IG triggering negative emotions among tweens and impacting their mental well-being (and) our ranking algorithms taking [them] into negative spirals & feedback loops that are hard t…
1excerpt
Messages & email · 3 pp
Miki Rothschild: "[W]ith the recent PR fire TikTok has seen around inappropriate content for teens (see here), we were chatting within Youth XFN about our own deficits when it comes to discoverability, namely: (1) it's still very easy to find borderline and so…
1excerpt
Research & report · 9 pp
"Reels featuring kids/tweens, and teens (i.e., 0-15) dancing and posing are pervasive, and often show up on people's Reels experience, even in cold starts. While much of this content is benign (they may be uninteresting to users, but not necessarily harmful), …
1excerpt
Research & report · 9 pp
"A lot of NSC has been attributed to celebrities and influencers, and we've currently been looking at controls and ranking to solve this." "However, it's not an influencer problem, it's a culture problem. Those influencers might be responding to Instagram's c…
2excerpts
Messages & email · 4 pp
Shilpa Mody: "[I]t turns out that at the high end, we really do recommend a LOT of high-NAC content to teens! Like for the top 10% of teens, our recs in Explore are 57% high-NAC :/"
1excerpt
Sources: published internal records, unsealed discovery exhibits and court filings in MDL 3047 and related litigation. An allegation in a filing is not an independently verified fact.
Meta & Children's Online Safety