Twitter's crowdsourced fact-checking program Birdwatch remains a small pilot project 13 months after launch, with only 359 contributors flagging tweets in 2022
With the Ukraine war unfolding on social media, parsing fact from fiction has never been trickier — or, for those involved, more urgent. Tweets: @willoremus , @kcoleman , @shannonpareil , @carnage4life , @willoremus , @mantzarlis , @kcoleman , and @blackamazon Tweets: Will Oremus / @willoremus : Over a year ago, Twitter launched “Birdwatch,” a crowdsourced fact-checking platform for tweets. Today, as Ukraine misinfo runs rampant, Birdwatch is still in beta, invisible to ordinary users, and forgotten by most. My story: https://www.washingtonpost.com/ ... w/ @jeremybmerrill @kcoleman : @WillOremus @jeremybmerrill We'll be expanding the pilot very soon. It's important to us that notes are helpful to people (incl ppl from a wide range of views) and we've been focused on making that a reality before expanding. It's essential to the product having the impact we believe it can. Shannon Bond / @shannonpareil : “For the most part, the fact-checking notes rated “helpful” actually did seem potentially helpful — that is, if they were incorporated into Twitter in any meaningful way, which they aren't.” https://twitter.com/... @carnage4life : Asking users to fact check tweets quickly becomes a Who Watches the Watchmen? problem. The viral tweets around the invasion of Ukraine are a good example as thousands of users endorse & retweet content few of them are even capable of vetting while in parallel bots post propaganda https://twitter.com/... Will Oremus / @willoremus : Twitter VP @kcoleman says the company will be expanding its “Birdwatch” crowdsourced fact-checking project soon, and the company wants to get it right before giving it more of a role on the wider platform. 👇 https://twitter.com/... Alexios / @mantzarlis : Haven't checked in on Birdwatch in a while, glad Will has. These days TW has been essential to my understanding of the war so I think that there is potential, even if it does not appear tapped yet. Part of the solution is motivation but part of it, I think, might be curation https://twitter.com/... @kcoleman : @WillOremus @jeremybmerrill There's no roadmap for a product like this, so we have to figure it out as we go. We're fortunate to have some amazing contributors in the pilot, and academic advisors helping us. We'll be sharing some incredibly novel work soon - stay tuned. @blackamazon : Because community management is a thing and especially for fact checking, but that was them trying to activate the intrinsic dynamics of Twitter communities... based on research that doesn't actually respect it https://twitter.com/...
Context & Ripple Effects
Birdwatch launched in January 2021 as a crowdsourced fact-checking pilot of roughly 1,000 users, but an early analysis found partisan rhetoric and thin source citations, and when notes rolled out across iOS, Android, and desktop they stayed visible only to pilot participants. Thirteen months in, the Washington Post reports just 359 contributors actually flagged tweets in 2022 — leaving the tool invisible to ordinary users precisely as Ukraine war misinformation spreads.
The reporting landed at an inflection point: the day after publication, Twitter committed to opening Birdwatch to a small, randomized set of US users, and by fall it had made notes visible to all US users while adding 1,000 contributors weekly ahead of the midterms.
First-order effects
- During the Ukraine information crisis, ordinary Twitter users see no Birdwatch context at all — the entire fact-checking burden rests on roughly 359 active contributors inside a closed beta.
- Twitter's own product leadership (@kcoleman among those quoted) faces immediate pressure to justify why a year-old anti-misinformation tool remains hidden from the feed where misinformation circulates.
Second-order effects
- The forced response is structural: expanding to randomized US users and then all US users converts Birdwatch from an experiment into production infrastructure, making note quality — the early partisan-rhetoric problem — a public-facing risk rather than an internal one.
- Scaling via 1,000 new contributors per week with 'rating impact' scores ties fact-checking throughput to volunteer incentive design instead of Twitter's moderation headcount, shifting cost but also control out of the company's hands.
Third-order effects
- If the pattern holds, major platforms converge on crowdsourced community notes as their default trust layer — with the open question being whether contributor-scale governance can fix the partisan skew documented since the pilot's first months.
- A slow, crisis-exposed rollout like Birdwatch's becomes the case study regulators and rivals weigh when deciding whether volunteer fact-checking can substitute for professional moderation at wartime speed.
The trend: Social platforms are migrating trust-and-safety from centralized professional moderation toward crowdsourced community fact-checking, with Twitter's Birdwatch the earliest large-scale test of whether volunteers can scale fast enough to matter.