████: "Hi ████! Wanted to raise something with you--was going to post this as a reply on Pratiti's post sharing Nick's letter, but decided to provide the feedback privately instead. I wonder if this is something we can raise with our comms team (or whoever writes those public-facing statements).
While I don't disagree with the intent or accuracy of the text from Nick's letter (snippet below), I wonder if we could work to be more specific going forward in these types of responses? Pre-FB I spent a lot of time working on public health and environmental issues, and this sounds eerily similar to what tobacco companies and climate change deniers say. Uncertainty/doubt is a key component of the scientific method, but it can also be weaponized to push back on critics (e.g., '...but this one scientist thinks cigarettes don't cause cancer,' 'we need more research to know for sure whether climate change is man made,' 'evolution is just a theory,' etc etc). I don't think that's what Nick was trying to do, but when we use language like this it puts us in very bad company....
'The truth is that research into the impact social media has on people is still relatively nascent and evolving, and social media itself is changing rapidly. Some researchers argue that we need more evidence to understand social media's impact on people. Each study has limitations and caveats, so no single study is going to be conclusive.'"
████: "thanks so much for the feedback Kyle! Can you share an example of what you mean by 'work to be more specific'?"
████: "Like, providing a synopsis of what we think our internal research says, and what areas legitimately need more research."
████: "I could imagine a one paragraph summary of Daisy--teens say IG makes certain aspects of well-being worse, but when we run experimental tests of different features, etc., we don't see stat sig changes in well-being measures, indicating to us that we can't point to anything specific in the product, etc"
████: "gotcha. I get your point. I don't know that we feel certain about what the internal research says but I do get your point :)"
████: "Yeah lots of legit disagreement internally in how to interpret the data. I know we wouldn't do this, but would be fascinating to sponsor a point-counterpoint newsroom articles. Evidence for and against the idea that IG is bad for teens"
████: "Our research is more nuanced and shows both good and challenging experiences for different people at diff times. I'll share the feedback if that's ok (prob more to Kristin and comms for future). Interestingly I didn't see nick's post before it went out"
████: "heads up, team - mark read the Ryan Mac NYT story, specifically the reference to how he posted a video this summer about riding an "electric surfboard". mark is eager to post the following:
'Look, it's one thing for journalists to make false claims about my work, but it's crossing a line to say I'm riding an electric surfboard when it's clearly a hydrofoil and I'm pumping that thing with my legs.'"
Nick Clegg: "Am I missing something here? On the day a [PP] rep is pulled apart by US senators on whether we care enough about children on our services, MZ is going to post about... surfboards? Maybe I've lost my humor about this whole thing, but I really think this would seem to any casual observer to be pretty tone deaf given the gravity of the things we're being accused of."
Nick Clegg: "the IG/teens story is huge, is cutting through, and will lead to countless testimonies in the weeks/months ahead. If I was him, I wouldn't want to be asked 'while your company was being accused of aiding and abetting teenage suicide why was your only public pronouncement a post about surfing?'"
████: "OK - message sent (and Ted clip) - will let you know what he comes back with"
████: "His response: 'I'm still going to post and attach that screenshot I sent you'"
Nick Clegg: "Fair enough -- I guess if it's consistent with stuff he's been doing every day I might be overreacting. I just worry that he and SS are under estimating how much this IG/kids thing will continue to hit us"
Nick Clegg: "TBH this isn't so much a daily comms coverage issue, it's more that the WSJ/IG story has dramatically consolidated a wider narrative (that we're bad for kids) which had been brewing for some time. It now makes regulation - first in the UK, EU and then elsewhere - certain, and in my view makes launching IG Kids nigh impossible. I've told MZ and SS this already."
Nick Clegg: "Hey - to state the flamingly obvious, the WSJ story about IG and teenage depression and suicide will have a huge impact on regulatory/political pressure on us going forward. Whilst this trend - focus on the impact of our services on kids - was already growing, it has intensified exponentially with the WSJ series. I'm worried that none of this is - yet - being reflected in Mark's decision making on head count etc. Can you remind me of the delta between the eng napkin allocation for youth and the initial bid?"
Susan Li: "I think the original bid was 140"
Susan Li: "And 50 were received here"
Nick Clegg: "Got it - and do you know by any chance what is likely to be dropped/cut back without the 90 heads?"
Susan Li: "I am not dialed in enough on Youth specifically to know, unfortunately"
Nick Clegg: "Understood - thanks - I don't want to spin up unnecessary work but the external environment has hardened more than I think Mark appreciated on this issue so I think it would be useful to quickly see what the gaps are"
Susan Li: "They had 80 ENG on the team and 62 unfilled heads from their 2021 Budget of 142 so just filling their existing allotment would represent tremendous growth for them"
Nick Clegg: "Hey – just wondering how we can follow up on the conversation in SG this morning re the case for further wellbeing research? It seems to me we need to significantly accelerate our existing plans – and ideally design a FORT style sharing of underlying data with relevant researchers asap. Mark seems open to us doing this now so we should move quickly before the winds change."
Pratiti Raychoudhury: "Yes, I can. From what I took away was – (1) move with urgency and be more prepared/ proactive (2) because well-being is such a vast topic, start with mental health and teens. A few thoughts that I have:
• For external research, we need to make a decision on data sharing quickly (given the concerns from IR). We have already identified that the longitudinal correlative study and the millennium cohort studies are the 2 best ones to start with. Both OII and us are ready to go as soon as we make a decision. I'll get this group and Jen + Mike Johnson together.
• We are also planning to do a more RFP style approach, similar to US2020. We are cataloging all the data needed for this and suspect will run into similar concerns from IR and legal. (These data sets will be more complex than the ones we want to share with OII).
• Look for 'special' collaborations – similar to what we did with Raj Chetty and social capital
• Red-team all existing research we have internally on mental health and teens. In the event of leaks, we can respond much faster.
• Create a one-pager (similar to the Polarization and Bullying and Harassment ones). This will provide a holistic view (internally) on teens and mental health since so far our response is based only on the leaked studies."
Nick Clegg: "Once we have nailed down some more of the specifics here, it would be good to produce a note for Mark setting out these plans and checking he agrees with the level of ambition and speed – especially in view of the skepticism of some FB Board members expressed today, which I understand will be the subject of a further conversation."
Adam Mosseri: "FYI I'm really worried about this. It's going to be incredibly hard to get done. We don't have full board alignment and the privacy and legal hurdles are significant. We've been talking about this for a long time but have made little progress. Would love your take on how to make this happen. I'd like us to be able to talk publicly about our plans by early December, and get something started in H1."
Emily Dalton Smith: "Got it – I understand your concerns about the difficulty and risk here, and am aligned.
I think we've made good progress on research and have identified the initial focus areas (problematic use, connections, bullying & harassment, and emotional impact). Our biggest gap is getting this research into product roadmaps. We got 0 new Well-being funding for 2022, so we'll need to tee up the tradeoffs against other priorities."
Adam Mosseri: "That's far too slow. I'm likely testifying in front of congress the first week in December, and today Snap, TikTok, and YouTube all committed publicly to sharing data with congress. Happy to jump on a VC if that's easier."
Kang-Xing Jin: "I think the key thing is we're being asked / pushed to do a bunch of stuff without the staffing right now, and so the sooner we can tee that up for unblocking and decision the better. On product, if we can resolve the 10 XFN thing (and then have process to get FoA to re-pri some work on things like problematic use), then we'll be more on track. On data sharing / research, I'm less sure what is needed (or specific goals / timelines) since I haven't been tracking as closely. But agree having a solid PM just on that would also probably be important – in addition to the research functions. And given the MZ well-being external research review I worked on for this topic in 2019, I 100% agree we've made not a lot of progress here in the last 2 years, so do need a re-set of approach and strategy."
Emily Dalton Smith: "We've made a lot of progress on research and data sharing. We've not made a lot of progress on getting the research into product."
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 to exit from."
"They keep saying leaks lower morale… no, Adam the choices you guys make lowers morale"
"On one hand we have Miki and ████ literally forcing the product team to make stupid decisions (teens dont want meditation, show lava lamps; or you can have the break without it being opt in first). On the other hand we get discredited for doing the best work we could."
"They (or ████) are responsible for the Take a Break being opt in btw."
████: "The results of BEEF, the large end-of-half TRIPS-like survey are only being shared in private and select groups, to avoid leaks. Sad new world"
████: "They are being shared in this 66-person secret group. ████ and ████ are in it, other let me know if you're interested? It's ... the 'IG Bad Experiences measurement working group'"
████: "Also, two full research decks with annotations just released publicly half an hour ago."
████: "And if you read the annotations, they are the worst. Whether or not the research was flawed, this makes me so upset."
████: "@silent i've said it before, and i'll say it again. the research is really great, and something to be proud of. i think it's better to have it all out there than just the parts WSJ decides are newsworthy. also, so even if you accept that 'Facebook knew!!!' the first appearance comparison note i see was posted in February. how quickly are we supposed to fix the problem? or are we just supposed to shut down the whole product?"
████: "@silent The media/Twitter narrative seems to be that either Instagram should just shut down, or that there's some magic wand you could use to eliminate negative social comparison. It's just a distraction, but I feel for the people involved (although I think Justin and Moira are taking it pretty chill)"
Shruti Bhutada: "So apparently FB published the deck before WSJ and thats really bad practice in comms"
Wendy Tegge Gross: "yup"
Shruti Bhutada: "I am more worried they keep making this into a bigger fire than hurt."
Shruti Bhutada: "And I am hella hurt"
Shruti Bhutada: "like PLEASE just focus on what they are REALLY criticizing you about and acknowledge it and stop throwing your researchers under the bus and double crossing journos so this can all end PLEASE"
Shilpa Mody: "There is a world where they could have annotated with actual context instead of a tear down, and that would have been a reasonable outcome"
Wendy Tegge Gross: "agree to all of this. I'm hearing it was direction from MZ"
Wendy Tegge Gross: "ended up being basically an MZ call on how it was handled"
Shruti Bhutada: "Like MZ asked Cox and Adam to say research used flowery language and asked researched leadership to as their own ICs to flame other ICs research?"
Wendy Tegge Gross: "I get the impression that MZ decided how to frame the external narrative, that other teams fought back hard and lost"
Shilpa Mody: "It does feel like the tone of the response has shifted... like Adam previously sounded like 'the wsj was misleading, and the numbers aren't as bad as they look, but also we acknowledge there's a problem and here's what we're doing about it'"
Shilpa Mody: "That last part has definitely dropped off now"
Wendy Tegge Gross: "For example: 'When reporting on qualitative research results, some conclusions/statements may be researcher speculation versus what was actually reported by participants' makes it sound like I was just guessing or making things up rather than finding the common themes across the qual groups."
Shruti Bhutada: "In the meantime, how do I respond to leaders checking in and asking how they can support me? Whats a non-causal, non-floral way of saying: You could start by not throwing us under the bus and giving us boiler plate replies?"
Shilpa Mody: "this is a good question! I just replied honestly to ████ and ████ about how I feel betrayed, that the town hall was tone deaf, etc. But not sure if that's also the right response to give to Kristin etc"
Shruti Bhutada: "Ok I wrote this to Diego, tell me if a version of this can be sent to Kristin etc: I am disappointed in senior leadership and how they are responding to this. I am disappointed in the constant use of vague words like 'floral, flowery, hyperbolic' to describe our words and the focus on how the research was done rather than the external critique which is about why nothing was done as a result of it. It was extremely hurtful and disappointing to see our work (and abilities) flamed in the annotations (and I dont know see that helps contextualize or critique the work in a useful way). But most of all I am just sad me, Shilpa, Wendy, Justin, you and ████ have to go through this."
Shilpa Mody: "Also.... if we feel like it, i think there could be an opportunity to reply to Adam's email and push him to be bolder on mwb work. Like literally the best thing you could do to support us right now is to take seriously the possibility that we haven't done enough here, and that we should consider bigger bets that take a stronger POV to show teens and the public that we actually care about this problem."
Wendy Tegge Gross: "It would need to be so diplomatic. Like support us by showing that you value the research and invest in the space. But not calling out that the company has thrown us under the bus and not making it seem like we don't understand trade offs the company has made"
Shruti Bhutada: "its a bit trippy to also be in some of the product threads where ████ or Miki will literally say, we should just do this, teens dont want that, what about lava lamps for take a break."
Shruti Bhutada: "I cant tell if this is some PR strategy where they publicly discredit their own research to save their backs"
Shilpa Mody: "It is absolutely a PR strategy to publicly discredit the research. I think they are seriously afraid of legal/regulatory repercussions if there are acknowledged causal links."
Shilpa Mody: "Btw did either of you watch any of the hearing?"
Shruti Bhutada: "One senator asked about our findings (looked like they had Moira and Justin's data from Dec 2020? or maybe Wendy? idk). And Antigone was like: The research is not a bombshell"
Shruti Bhutada: "and he was like: No, actually this research is a bombshell"
Shruti Bhutada: "Another asked about what IG did about the findings"
Shruti Bhutada: "And Antigone started saying its nor causal yada yada"
Shruti Bhutada: "And the senator was like: I am asking if you were aware and if you did something about it"
Wendy Tegge Gross: "it's not a bombshell internally but it is a bombshell that these internal research decks exist. So both the senator and Antigone are right"
Shruti Bhutada: "And there was more: research isnt causal and its not that bad yada yada"
Shruti Bhutada: "The senator was asking about what IG did so I think it wasnt helpful to just be like: research isnt causal. Since causality isnt the point."
Shruti Bhutada: "I... dont know how they get it this wrong"
Shruti Bhutada: "Like just answer their actual question. They know you think the research is shit but in this area theres also external research. And 1 in 3 or 1 in 5 makes no difference when they are asking you what you did about that 1."
Shruti Bhutada: "or maybe they cant say that. idk."
Shruti Bhutada: "I felt a bit bad for Antigone."
████: "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 groups next week."
"Summary
1. Generally, the countries which have enough samples to measure Teen prevalence rates have significantly higher SWOP in the teen segment than the general population.
2. Age-based prevalence measurement on Instagram has major recall gaps, but Teens in countries with measurement are seeing significantly higher integrity harms compared to the general population. We need to upsample this User segment that is a company focus and vulnerable population.
3. Teens see 2-4x SWOP in every country we have enough samples to measure except India. SWOP heavily weights Bullying and Harassment, which appears to be a relatively uniform and prevalent harm amongst Teens. Teens are also subject to higher rates of numerous other harms, including SSI and VNI. These harms tend to manifest on Hashtags and Comments."
"Recommendations
1. We need to Increase sampling for Teens in key markets once support for easier upsampling is possible.
2. Age classification is currently a limiter, and we should push strategies to improve recall of Teens. This could mean accelerating strategies to request age from existing accounts and/or revisiting our protocol to not share age with family matching.
3. Consider adopting strategies to reduce integrity Harm prevalence in the Teens user segment. SWOP reductions goals or counter-metrics should be considered in context to growth goals and experimentation. This could include showing age-appropriate content and/or adjusting ranking and recommendations based on age.
4. We should consider what levers could be applied or developed to reduce the harm experienced by users under 18, particularly for the areas of Bullying and Harassment + SSI. Given the connection of these harms to Hashtags and Comments, it aligns with other topline observations that those surfaces have higher rates of harm.
5. Push for improved IMI calculation methodology and tooling to make Teens segmentation available to end users in dashboard formats."
████: "Additional TLDR thoughts:
1/ SWOP prevelance for teens is ~0.42% of the IG ecosystem. This might seem like a small number but its still ~2x worse for teens than non-teens (0.24% vs 0.42%). It also appears to be especially bad for hashtags and comments (suggesting product opportunities around discoverability/search and proactive close friend/friend interactions).
2/ We also ran a very preliminary labeling exercise where we found that ~30% of labeled content appeared to be bad or borderline 'icky' things we would not want kids to be exposed to. We only looked at 50 pieces of content and did not use policy guidelines to do the labeling; post forthcoming from Aashna but this should give us some level of intuition that the problem is much larger than the 0.42% of things we are currently capable of measuring and actioning against with current policy.
3. 1/50 pieces of content was 'age appropriate nutritious' or the sort of content we would like to promote to teens. This should also suggest that we have a big hill to climb on ensuring a higher % of a teen content consumption diet is 'nutritious'.
4. So... we have a big challenge both with closing gaps against conventional integrety defences but also with developing new policy, measurement, detection, parental controls, and enforcement (in the form of both removal and soft actions)."
"51% of Instagram users say 'yes' to having had a bad or harmful experience in the last 7 days. Out of those 1% of report and of those 2% have the content taken down (i.e. 0.02%). The numbers are probably similar on Facebook."
"21.8% of 13-15 year olds said they were the target of bullying.
39.4% of 13-15 year olds said they experienced negative comparison.
24.4% of 13-15 year old responded said they received unwanted advances"
"I believe that it is important to get the following efforts well-funded and prioritized."
Bullying & harassmentSuicide & self-injuryChild sexual exploitationHashtags+1
"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 types.
2. ~12% of content potentially may be "age inappropriate" for teens but not covered by current policies.
3. We estimate 1 in 10 pieces of content (~9%) may be 'nutritious'"
"Teens see >2x prevalence for B&H, VNI, and SSI."
"The drivers of higher SWOP are Bullying and Harassment (B&H), Suicide and Self Injury (SSI), and Violence and Incitement (V&I), which are all >2x higher prevalence for teens than the average user."
"Teens see ~9x SWOP in Hashtags compared to the average User; particularly across SSPx, Sexual Solicitation, and Nudity"
"Comments see high rates of Bullying and Harassment for Teens"
"Teens also see higher prevalence of harms on Home, particularly B+H, SSI, and VNI"
"While 0.42% of content is policy violating, we have reason to believe that ~12% of content is potentially inappropriate for early teens. And after viewing or engaging with inappropriate content, it starts to show up more and more on recommendation surfaces throughout the app."
"And of the ~88% of content that we believe may be appropriate for teens, only ~1 in 10 of them may be 'nutritious' for teens."
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 tuning on a number of integrity areas to correct for over/under-enforcement (including body positivity?? which he used as what felt like defense against NSC), comments integrity, age-related recommendation tuning, and rabbitholing engineering on the back burner because we're so deeply understaffed"
Alison Lee: "because our p0 as a team is just to keep basic integrity and non-rec content from surging and foundational engineering improvements, and we're 4 HCs short (and really, we need 8 more engineers and at least 1 more researcher)"
Alison Lee: "so to hear adam wonder out loud, 'how much is ever enough?' was frankly a huge disappointment"
Lori ████: "I'm not sure I follow this last pt. Are you talking about your specific team or something broader?"
Alison Lee: "yes, relevance integrity team"
Lori ████: "I understand that. It does feel like a cop out to go to the other extreme and say we could put everyone on this and that wouldn't be the right move."
Lori ████: "there are clearly lots of alternatives that could be more helpful to discuss in between the extremes."
Alison Lee: "tbh we should actually put everyone on this, because safe experiences = better engagement, no? more equitable experiences, safer, more inclusive would = better engagement. to position resources as 'integrity/wellbeing vs. engagement' feels like a false dichotomy here"
Alison Lee: "and part of the reason why we're losing out to tiktok, as well as with young people"
Alison Lee: "or at minimum, unsafe experiences = decreased engagement"
"I care deeply about this and we'll continue focusing on being industry-leading with young people, but I'm not going to say [child safety] is my personal main focus when I have a number of other areas I'm more focused on like building the metaverse."
████: "in the past, the only requirement for age-gating on IG was by advertisers, around brand safety… That's why the relevant infra and code are owned by the IG Ads team ('IG4B') as opposed to Well-being. We never used that infra for Integrity. I don't even know how reliable it is… Not opposed to enhancing it, just sharing that it was never designed or used for Integrity purposes in the past."
████: "It sounds like some folx just assumed age-gating would work for Integrity, when it was never planned to."
████: "It wasn't a 'gap' until now because nothing depended on it or tried to use eit (outside the well-known Ads use-cases)."
████: "This has been part of policy for years (pre CI taking on IG directly, even), so the better question may be why IGWB never built this"
████: "This is literally the first time I hear about this infra being used for Integrity purposes, even in theory, since I've worked in the domain (2018)."
████: "IGWB handed this off to Ads years ago because we weren't using it at all, much less building it"
████: "The gap is external community standards pages mentioning we age gate but not implemented in IG product."
████: "There's a bunch of work, spurred on by the Youth work, that is helping to improve coverage, but we didn't even ask any IG users for age until a few months ago."
████: "I think Adam knows what this infra is used for, or not. A heads-up to Adam is OK… but it won't be news to him"
"Sheryl,
Please find the updated Cheat Sheet based on your feedback today:
• Added the toplines you wanted from Monika and Antigone's prep docs
• Updated with latest figures on hate prevalence and % of hate we remove before seen vs what it was in 2017 (first started reporting on this)
• Included the default to private accounts for teens
• Did not include 'Hiding Likes' even though you asked for it. This is because it was something that WSJ cited. Leaked research showed that 'Hiding Likes' had no effect on wellbeing, but we did it anyway for PR.
• The included examples of IG features to help with teen wellbeing and body image are the ones that the IG team said are the best to use -- they used these examples for Antigone's Congressional testimony.
◦ They are broken down into two categories: friction and interruption
Please let me know if you want me to make any other additions and changes."
"Have you ever seen anyone discriminating against people on Instagram because their gender, religion, race, sexual orientation, or another part of their identity?
13-15 year olds: 26%/31%"
"The launch criteria are… 3. Balance ecosystem effects: The balance between the known upside of the solution and changes to ecosystem usage metrics. We'd expect some level of effect if there's material usage."
"We recommend prioritizing our analysis of experiment data on #3"
"For #3, we will:
- Disregard variants with material ecosystem regressions:
- Time spent: regressions < 0.2%
- Sessions: regressions: < 0.2%
- Creator watch time: regressions < 0.5%
- Balance ecosystem effects:
- If there are multiple variants at this point, then our choice of the best variant prioritizes trading off big reductions in ecosystem regressions for small reductions in adoption."
"I think what we're saying is that we want to guard against any ecosystem/engagement metrics that imply we're hurting the user experience and that we'll rely on research to indicate the upside, but not hold ourselves to measurable upside (improvements in user reported feelings of NAC) since we don't yet have a great way to measure that in a test setting."
████: "so, there appears to be a team of lawyers scheduling meetings with individual researchers to question them about well-being research. ████ has already done a one-hour meeting and it sounds like they want more. i just became aware of this, and i'm afraid they are going to ask for a lot more of your time. i'm trying to head them off and push for something that is more efficient, like a group presentation."
████: "meanwhile in the neverending review cycle on our internal note, our comms person has now added two more comms people to review before we can post internally to closed groups."
████: "Yep. ████ asked for Stephanie Otway (her manager) and ████ (another comms person). ████ said: 'i also want to share with abby and stephanie otway for overall gut check. im so sorry to slow this down but i really want to make sure this doesnt get shared in a way that harms the team's good work so if you can give us a couple of days to work through on the comms side, that would be much appreciated'"
████: "But (and I didn't tell her this), I think giving others a heads up is reasonable, but not gating the document on a gut check from comms."
████: "Ok cool. I totally agree with you. I was thinking a lot about this as I lay (jetlagged) in bed early this morning. The angle I'm thinking about, especially when comms is gating, is how can we share research findings that are inconvenient for facebook, e.g., research that shows fb is harmful. we've accepted the complexities of sharing such research externally, but if it can't be done or it needs to be sanitized to share with people that need to know (i.e., the people in focused, closed groups) then we've got a big problem."
████: "Talking with ████ yesterday, he mentioned a comment from some 'reviewer' on one of the note. They took issue with language describing a finding as applying to a general population instead of just survey responders. He explained this is why we weight survey responses. The discussion that followed left him feeling that they wouldn't want us to do that so that we could more easily dismiss inconvenient findings. This is a huge moral hazard, in my opinion."
████: "agreed!"
"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, which introduces risk:"
"Product/Research Risks [KX/Emily, Pratiti]: Each platform prioritizes different topics and uses different definitions of success without clear rationalization. We end up with duplicative and contradictory research and product at times (e.g., IG research giving different recommendations for topic prioritization than Social Impact/DSS research). Lack of centralization akin to privacy or integrity means topics may be under-represented in prioritization at exec levels."
"Policy Risks [Antigone]: Policymakers and stakeholders consistently press for evidence of senior leadership, a dedicated team and significant investment and offer solutions like criminal liability (e.g., director level criminal liability under consideration in UK Online Harms Bill). Policymakers and stakeholders see the lack of a consistent, comprehensive approach to problematic use and certain 'harmful content' (e.g., SSI + ED content) across FOA as evidence we prioritize engagement/profits; they are offering solutions aimed at algorithmic personalization, which they believe drives 'addictive' engagement (e.g., 'Justice Against Malicious Algorithms'). Our Youth XFN is not resourced to take an FOA approach, nor to rapidly address existing privacy, safety and well-being gaps simultaneously, meaning we will not meet policymaker expectations until end of 2022. The slower progress could fuel support for the more draconian measures in numerous online safety bills making their way through the legislative process globally."
"Comms/Marketing Risks ████: There is no proactive strategy for how we talk about our work on these topics outside of IG (e.g., when the WSJ leaks happened, Comms piecemealed material). Our last major company announcement was time controls in Aug 2018. There is a campaign being run against us and well-being is a key element. Well-being is also a core narrative/ATW that marketing research shows is a top 3 priority topic that could hold us back from being trusted to help build the metaverse."
"Partnerships Risks ████: We're at risk of losing our public facing partners. Our lack of action has been noticed to the point of creating internal challenges for our partner POCs, and a lack of trust that we are acting on recommendations that they have facilitated for us in roundtables."
"A central team would give us consistency on strategy, process and quality because these issues will continue otherwise. We have already seen early success with this model for Health and Equity."
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