Meta & Children's Online SafetyChat: Somewhere between zero and negligible. Child safety is an explicit non-goal this half.
Raw text
Case 4:22-md-03047-YGR Document 2758-36 Filed 02/20/26 Page 1 of 11 AMENDED Exhibit 378 PLAINTIFFS' OMNIBUS OPPOSITION TO DEFENDANTS' MOTIONS FOR SUMMARY JUDGMENT Case No.: 4:22-md-03047-YGR MDL No. 3047 In Re: Social Media Adolescent Addiction/Personal Injury Products Liability Litigation Case 4:22-md-03047-YGR Document 2758-36 Filed 02/20/26 Page 2 of 11 Message From: [/O=THEFACEBOOK/OU=EXTERNAL (FYDIBOHF25SPDLT)/CN=RECIPIENTS/CN= Sent: To: 7/10/2020 12:50:48 AM 1; Miki Rothschild 1 Subject: Message summary [{"otherUserFbId":null,"threadFbld":2197046077008938}] (7/09/2020 08:08:11 PDT): >Good morning! FYI my xfinity internet has been down all night and it was flickering off and crappy yesterday. So I may be just calling in to our reviews (7/09/2020 08:23:41 PDT): >I've done a couple walking around outside on phone. It works great! 100 (7/09/2020 08:35:15 PDT): >I tried a few of those early on in shelter but what's challenging is that the cell signal is also not strong around my neighborhood (7/09/2020 08:35:33 PDT): >So it made 1:1s challenging : ( (7/09/2020 08:37:40 PDT): >Time for one of those 1970s era phones with super long cord? Or maybe a 1980s brick+antenna combo? ;) (7/09/2020 08:38:11 PDT): shared: 62211832_481162075964625_4604683761446027264_n.gif (7/09/2020 08:59:27 PDT): >Good morning leads! What will the day bring today? :) (7/09/2020 08:59:47 PDT): >what is the plan for the Monthly check-in with Comm. Leads? (7/09/2020 09:00:00 PDT): >@Miki shared a first draft doc via email last night. (7/09/2020 09:00:08 PDT): >I'm looking at it right now :) (7/09/2020 09:00:29 PDT): >https://fb.quip.com/H5GsAabcS8mk (7/09/2020 09:03:55 PDT): >Well, on the plus side, I guess our goals are aggressive enough :) Lots more missed this half. (7/09/2020 09:18:02 PDT): >I like the on-track/off-track grid. (7/09/2020 09:22:28 PDT): >Hey, folks! Just FYI that though is still in boot camp, she may be attending the check-in with Community leads today. The H1 retro and H2 lookahead is hard to pass up. (7/09/2020 09:25:58 PDT): >How long is her Bootcamp? (7/09/2020 09:26:54 PDT): >It's supposed to be two weeks but she's steamrolling through a lot of the (mostly self-driven) content. She's likely going to start poking in early next week. (7/09/2020 09:47:58 PDT): >Just FYI, I have to drop off the Bullying mtg at the halfway mark to make a 10:30 pxfn review (7/09/2020 09:51:45 PDT): >I also will need to drop off halfway through for a wb measurement review w/ EDS KX and John hegeman (7/09/2020 09:51:52 PDT): >At 10:30 (7/09/2020 09:52:26 PDT): HIGHLY CONFIDENTIAL (COMPETITOR) META3047MDL_RVOL003 META3047MDL-003-00069904 Case 4:22-md-03047-YGR Document 2758-36 Filed 02/20/26 Page 3 of 11 >I'll be a few min late to this meeting running between leadership candidate debriefs. (7/09/2020 09:52:51 PDT): >err, the bullying session. (7/09/2020 09:54:00 PDT): >But I am very excited for this meeting with our CDS partners. Excited for y'all to meet them and hear some of the great work they have been doing and see and share any feedback on some of the plans for the half (7/09/2020 10:01:09 PDT): >getting error on bjn (7/09/2020 10:01:13 PDT): >████ (7/09/2020 10:01:17 PDT): >we cannot dial in (7/09/2020 10:01:32 PDT): >samsies (7/09/2020 10:01:44 PDT): >Oh no! (7/09/2020 10:02:14 PDT): >can you pls update with zoom? (7/09/2020 10:03:29 PDT): >Just updated (7/09/2020 10:03:30 PDT): >To join the meeting on a computer or mobile phone: https://bluejeans.com/590196758?src=calendarLink > > > >One-Touch: +14087407256,,,,590196758#,,# > >Meeting ID: 590196758 >To join via phone: >1) Dial: +1-408-740-7256 >2) Enter Conference ID: 590196758# > >Want to test your video connection? >https://bluejeans.com/111 (7/09/2020 10:03:46 PDT): >I used BJN because the last time we did Zoom people did not like you couldn't join from a portal (7/09/2020 10:03:53 PDT): >Happy to do either (7/09/2020 10:04:18 PDT): >still erroring (7/09/2020 10:04:25 PDT): >works for me now (7/09/2020 10:04:34 PDT): >https://bluejeans.com/590196758 (7/09/2020 10:05:00 PDT): >Thanks a lot, ! (7/09/2020 10:05:13 PDT): >it booted me out via phone when I joined. (7/09/2020 10:05:31 PDT): >https://bluejeans.com/590196758 (7/09/2020 10:05:35 PDT): >working now (7/09/2020 10:11:20 PDT): >is it working for you? (7/09/2020 10:11:55 PDT): HIGHLY CONFIDENTIAL (COMPETITOR) META3047MDL_RVOL003 META3047MDL-003-00069905 Case 4:22-md-03047-YGR Document 2758-36 Filed 02/20/26 Page 4 of 11 >the phone number did not so I went through app and just turned off video to reduce bandwidth for me. but it is working (7/09/2020 10:16:45 PDT): >fyi - Core Data Science is a very expensive resource the interview process is more akin to ML/eng interview than a product DS. These are highly capable individuals and they only work on WHAT THEY WANT TO work on (managers don't allocate). We need to keep them excited and committed to IG WB. Challenge will always be relating work to product strategies. (7/09/2020 10:19:17 PDT): >that latter part, making sure their work informs product, is what has been especially compelling for many of them to work with us since in the past when they've worked on WB work on fb app/central, it wasn't leading to any product changes. So thoughts on what would be helpful in shaping their research to inform product is esp great here. Michael Yehuda Rothschild (7/09/2020 10:21:04 PDT): >BTW, this session is exactly what I was envisioning for our "Understand Series" (7/09/2020 10:21:17 PDT): >this is actually the first of that (7/09/2020 10:21:58 PDT): >we had been trying to get it scheduled as part of that but eventually just put in in our review slots. I think the idea is to continue to use the review slots moving forward, correct? Michael Yehuda Rothschild (7/09/2020 10:22:17 PDT): >In my old team we'd do these every week (rotating on product area) starting with 15 minutes on health/goals reporting update and then 45 minutes deep dive Michael Yehuda Rothschild (7/09/2020 10:28:37 PDT): >Yes, I think that would be amazing. I think we said bi weekly ~30 minutes. (7/09/2020 10:31:28 PDT): >That seems relevant to our discussion about Ecosystem Safety's future :) (7/09/2020 10:32:29 PDT): >We need to be careful here with messaging to our team, if any :) Technically we don't tell them what to work on either, they choose. And we don't disclose levels, comp, etc. (7/09/2020 10:32:47 PDT): >(PD on bullying on this meeting) is really excited about this meeting and how it can impact her product thinking. Love this meeting. (7/09/2020 10:33:50 PDT): >not tied to bullying, but be great to think about how to reduce these types of comments as well...maybe through comment warnings (7/09/2020 10:33:52 PDT): shared: 107699969_3188874801155484_9066211590323271398_n.png (7/09/2020 10:47:44 PDT): >do we ever add strikes against actors using classifiers or can only co do that? (7/09/2020 10:48:17 PDT): >I imagine the accuracy of classifiers would need to be very high (7/09/2020 10:56:17 PDT): >We (Facebook Inc) add strikes against actors using classifiers all the time. Many thousands per day. (7/09/2020 10:56:23 PDT): >CO only does manual enforcement. (7/09/2020 10:56:46 PDT): >Right, for strikes. But the nice thing about comment warnings (your idea above) is that it's not a strike. It's a nudge to the user, which they can ignore. Which means we can (and do) run the same classifiers at lower precision. (7/09/2020 10:57:38 PDT): >Love the ideas of warnings and comment mgmt coming from the classifiers! (7/09/2020 10:57:53 PDT): >That's the best leverage of the in-product nudge-type interventions. They're not enforcement. We've been using the same classifiers in lower precision for comment and post warnings for many months now :) (7/09/2020 10:58:12 PDT): >I think we launched comment warning (using the classifiers) ~12 months ago :) Nice early bullying win. HIGHLY CONFIDENTIAL (COMPETITOR) META3047MDL_RVOL003 META3047MDL-003-00069906 Case 4:22-md-03047-YGR Document 2758-36 Filed 02/20/26 Page 5 of 11 (7/09/2020 10:58:21 PDT): >CIFM, of course, complained about potential censorship, nanny state, etc. (7/09/2020 10:58:40 PDT): >Thankfully we ignored that criticism, launched it, it's worked well, and the following half (2019H2) the CI bullying and other teams started doing some of the same :) (7/09/2020 10:59:31 PDT): >This is one of the examples I use in engineer, EM onboarding, of why parity is the right goal but consistency isn't. (7/09/2020 10:59:44 PDT): >I used to use Daisy as well but we as a company have managed to screw that one up : ( (7/09/2020 10:59:46 PDT): >right - but having classifiers related to user perceptions is slightly different and potentially applicable to a broader user base (7/09/2020 11:01:52 PDT): >How so? (7/09/2020 11:02:19 PDT): >The comment warning classifier is literally "some people might find this offensive" and applied to every single user... (7/09/2020 11:03:54 PDT): >Current classifier is trained on violating labels.. the user perception classifier will beyond violating + low quality. Warning language should likely change (7/09/2020 11:04:35 PDT): >but I agree @ these are incremental improvements to the existing warnings Michael Yehuda Rothschild (7/09/2020 11:17:04 PDT): > > does this note answer your question about precise definition of problem maturity? https://fb.workplace.com/notes/nischal-pathania/content-integrity-problem-team-maturity- plan/383086065949597/ (7/09/2020 11:17:06 PDT): >The bullying comment classifier is trained on a broader policy definition, not the Community Guidelines. Lots of border-line, aggression, non-violating samples included in training (intentionally). Michael Yehuda Rothschild (7/09/2020 11:18:49 PDT): >And this is the complementary surface maturity definition: https://fb.workplace.com/notes/sam- england/what-is-the-role-of-a-surface-integrity-team/987352534949961/ (7/09/2020 11:18:51 PDT): >The classifier docs (https://www.internalfb.com/intern/wiki/Instagram_Bullying_Engineering_Team/aggression/Bullying/text_filt er_summary/text_classifier/policies_labeling/) link to the specific policy, with examples. (7/09/2020 11:32:08 PDT): >did I see on the line? (7/09/2020 11:32:34 PDT): >Is this the same as s counterpart on CI? or do we have a Vlad on foundation? (7/09/2020 11:33:08 PDT): >nm, i have my answer (7/09/2020 11:36:39 PDT): >The org pairings in CI are (eng) <> (product). reports to pairs with *and* from product, who both report to So it's not exactly 1:1 even in pairs, lol :) (7/09/2020 11:36:58 PDT): >is the problem team eng lead. other reports are (CI Foundation) and (Civic). (7/09/2020 11:37:54 PDT): >What would I do without you diligently setting me straight every time :) ? (7/09/2020 11:40:26 PDT): >Very sorry to be missing Reels sync; my meeting's running over (7/09/2020 11:41:30 PDT): >lol not every time, only when you're asking in areas where I already made mistakes :) (7/09/2020 11:42:10 PDT): HIGHLY CONFIDENTIAL (COMPETITOR) META3047MDL_RVOL003 META3047MDL-003-00069907 Case 4:22-md-03047-YGR Document 2758-36 Filed 02/20/26 Page 6 of 11 >Quote from (who's in the meeting now, awesome guy) last time I asked him about CI org structure: "honestly, I wish we'd called it CI MPK, CI LON, CI SEA. Nothing else passes the sanity bar when you look at the current org structure." l01 (7/09/2020 11:42:23 PDT): >crap... so there are many more areas where I need correction who is taking care of those? (7/09/2020 11:42:58 PDT): >what specifically are we doing for child grooming (something I just heard about that is happening a lot on TikTok)? (7/09/2020 11:43:36 PDT): >Somewhere between zero and negligible. Child safety is an explicit non-goal this half. I'd argue we're making it worse with Interop, but's that's a can of worms. (7/09/2020 11:43:48 PDT): >CI has the usual CEI and IIC teams working to improve classifiers. I'll share their roadmaps. (7/09/2020 11:43:56 PDT): >So we'll get some benefits without doing any work, hopefully. (7/09/2020 11:45:51 PDT): >Here you go, https://fb.quip.com/Cj8aAejuMsD7. (7/09/2020 11:46:07 PDT): >Get IG to 2-measured for "tier 1" IIC. Nice win if we get it. (7/09/2020 11:46:28 PDT): >We've punted on our teams investing in these areas for awhile as part of prioritization and because we have central teams working on these that as Yoav said, we hope* we will benefit from. But, several of the extreme harms areas have always seemed like something we could do more to focus on and support/partner w/ central given IG challenges specifically. But, it's always a question of ruthless prioritization and partnership models. (7/09/2020 11:47:05 PDT): >I'm particularly thinking of SSI, CEI, IIC... (7/09/2020 11:49:18 PDT): >I gotta say part of the thing that makes me most anxious about Reels is that it gives me a little PTSD around when we launched TBO (Explore v2) and Trending hashtags at the top of Explore. (7/09/2020 11:50:19 PDT): >thankfully, our integrity processes have matured a lot since then, but it was a mess having so much bad content very quickly getting crazy distribution on the top of explore for everyone. (7/09/2020 11:51:04 PDT): >non-rec measurement is actually in place for certain problem areas (7/09/2020 11:51:21 PDT): >guardrails have been set (7/09/2020 11:51:31 PDT): >we can argue however whether that's good enough (7/09/2020 12:06:14 PDT): >I flagged to Comm. Leads that the Reels call is running late (7/09/2020 12:06:26 PDT): >We are waiting for all to join (7/09/2020 12:06:51 PDT): >It's great that you ran over because we JUST finished with Growth. If anyone asks, we were not late either. (7/09/2020 12:54:10 PDT): >did you want to touch on stress tests (7/09/2020 12:56:42 PDT): >Not more thsan these topics. Thanks for asking :) (7/09/2020 13:09:42 PDT): >WB Design is interested in implementing this program for making sure our craft bar remains high. Would love to get your thoughts if this is something you think engineers would be supportive of? Katherine is going to head this up for the design team and will have designers start the conversations with their eng partners. https://fb.workplace.com/groups/IGDSFYI/permalink/705571776684434/ (7/09/2020 13:55:04 PDT): >[pages 7–11 fully redacted in filing] HIGHLY CONFIDENTIAL (COMPETITOR) META3047MDL_RVOL003 META3047MDL_RVOL003 META3047MDL_RVOL003 META3047MDL_RVOL003 META3047MDL_RVOL003 META3047MDL_RVOL003 META3047MDL-003-00069908 Case 4:22-md-03047-YGR Document 2758-36 Filed 02/20/26 Page 7 of 11 META3047MDL-003-00069909 Case 4:22-md-03047-YGR Document 2758-36 Filed 02/20/26 Page 8 of 11 META3047MDL-003-00069910 Case 4:22-md-03047-YGR Document 2758-36 Filed 02/20/26 Page 9 of 11 META3047MDL-003-00069911 Case 4:22.md-03047.VCR Document 2758-36 Filed 02/20/26 Page 10 of 11 Page META3047MDL-003-00069912 Case 4.22-md-03047-YGR Document 2758-36 Filed 02/20/26 Page 11 of 11 META3047MDL-003-00069913 HIGHLY CONFIDENTIAL (COMPETITOR) HIGHLY CONFIDENTIAL (COMPETITOR) HIGHLY CONFIDENTIAL (COMPETITOR) HIGHLY CONFIDENTIAL (COMPETITOR) HIGHLY CONFIDENTIAL (COMPETITOR)