Opens profile photo
Follow
Click to Follow yoyoel
Yoel Roth
@yoyoel
Former Head of Trust & Safety at
@Twitter
. PhD from
@AnnenbergPenn
.
@pastrami
's human. @yoyoel@macaw.social
San Francisco, CAyoyoel.comJoined May 2008
720 Following
200.1K Followers

Yoel Roth’s Tweets

Yoel Roth
@yoyoel
If there’s one takeaway, it’s this: What matters most in platform governance is how decisions get made. You can armchair quarterback specific choices and mistakes all day. But the real work is figuring out how to make principled decisions when all you have are bad options.
7,944
2,673
Show this thread
Yoel Roth
@yoyoel
Replying to
@yoyoel
Those steps take away many of the gratifications of the trolling here. Bans are frustrating and take work (and resources) to try to circumvent; visibility filtering deprives trolls of the attention and reach they're trying to get.
420
891
Yoel Roth
@yoyoel
Update on our efforts to combat hateful conduct: We've not only mitigated the recent surge in harmful behavior, but have reduced impressions on this content in Search by ~95% relative to even prior baseline levels. We're continuing our work to make Twitter safer every day.
Graph titled "Total impressions on Tweets with 1+ slur" showing a surge in impressions around October 28-30, followed by a decrease to levels below the prior baseline of impressions.
Quote Tweet
Yoel Roth
@yoyoel
Since Saturday, we’ve been focused on addressing the surge in hateful conduct on Twitter. We’ve made measurable progress, removing more than 1500 accounts and reducing impressions on this content to nearly zero. Here’s the latest on our work, and what’s next.
Show this thread
4,731
21.8K
Yoel Roth
@yoyoel
No solution to identity is perfect, and we’re iterating quickly to come up with the best approach here. We appreciate all the feedback, and will share more as our work progresses.
637
4,521
Show this thread
Yoel Roth
@yoyoel
It’s not just the $8, though — this kind of in-app purchase requires bad actors to get through not only Twitter’s defenses, but also those of mobile device manufacturers and payment processors. That’s not impossible for a sophisticated adversary — but it’s hard to do at scale.
252
4,018
Show this thread
Yoel Roth
@yoyoel
Proof-of-humanness is more straightforward. Most spam-fighting is a numbers game: Your objective is to raise the costs on spammers until their losses exceed their gains. When you do that, the spammers go away, and you win. $8 per account raises the costs of spam A LOT.
251
3,813
Show this thread
Yoel Roth
@yoyoel
For years, Verification on Twitter has been tricky because it’s a signal both of authenticity (you are who you say you are) and notability (you’re “important” by some standard). Notability is inherently difficult to determine in a fair way globally. I support getting rid of it.
181
4,257
Show this thread
Yoel Roth
@yoyoel
Long-term, I think we need to invest more in identity verification as a complement to proof-of-humanness. Paid Verification is a strong (not perfect) signal of humanness, which helps fight bots and spam. But that’s not the same thing as identity verification.
217
3,396
Show this thread
Yoel Roth
@yoyoel
We’re particularly focused on the risks of impersonation of public officials in the context of the US 2022 midterm elections, and have made the decision not to launch Verification for Twitter Blue until after Election Day.
94
3,617
Show this thread
Yoel Roth
@yoyoel
In the short term, we’ll ramp up proactive review of Blue Verified accounts that show signs of impersonating another user. When we find them, we’ll suspend them. See something that looks off? You can report it directly in the app.
91
2,933
Show this thread
Yoel Roth
@yoyoel
The planned changes to Twitter Blue to make Verification more widely available raise the stakes for this kind of impersonation. Here’s what we’re going to do about it:
48
2,746
Show this thread
Yoel Roth
@yoyoel
When Verified accounts use impersonation as a tactic — whether for parody or not — it creates an especially confusing experience. It’s been our long-standing practice to suspend Verified users when they do this.
131
4,759
Show this thread
Yoel Roth
@yoyoel
Verification! Impersonation! Twitter Blue! There’s a lot going on around identity on Twitter — let’s break down what our policies are, and some of the big questions we still need to answer…
1,695
25.7K
Show this thread
Yoel Roth
@yoyoel
I’ll continue to share updates about our Trust & Safety work as things evolve in the days to come. Our team's mission of enforcing our policies and protecting the conversations happening on Twitter remains unchanged.
247
3,371
Show this thread
Yoel Roth
@yoyoel
What AREN’T we doing? In the short term, we’ve had to deprioritize a few workflows — like account access (lost password requests), and some suspension appeals. We’re working to get these back online in the days to come.
204
3,281
Show this thread
Yoel Roth
@yoyoel
With early voting underway in the US, our efforts on election integrity — including harmful misinformation that can suppress the vote and combatting state-backed information operations — remain a top priority.
880
4,866
Show this thread
Yoel Roth
@yoyoel
More than 80% of our incoming content moderation volume was completely unaffected by this access change. The daily volume of moderation actions we take stayed steady through this period.
Image
188
3,791
Show this thread
Yoel Roth
@yoyoel
Last week, for security reasons, we restricted access to our internal tools for some users, including some members of my team. Most of the 2,000+ content moderators working on front-line review were not impacted, and access will be fully restored in the coming days.
137
3,798
Show this thread
Yoel Roth
@yoyoel
Yesterday’s reduction in force affected approximately 15% of our Trust & Safety organization (as opposed to approximately 50% cuts company-wide), with our front-line moderation staff experiencing the least impact.
236
4,484
Show this thread