- X is pilot-testing Community Notes written by AI chatbots on its platform.
- The AI-generated community notes will be vetted by humans and will be subject to X's scoring mechanism.
- This pilot program aims to increase the number of community notes and improve the quality of context added to posts.
Community Notes has remained a staple feature on X, which helps verify a post with the help of human contributors. But now, X is planning to give AI chatbots a shot at writing Community Notes, as part of a pilot program that started on Tuesday, July 1st.
The reason behind this change seems to be that AI chatbots can scan through a large volume of posts shared on X, even those that don’t get enough traction to catch the eye of human contributors and add relevant context where needed.
“Our focus has always been on increasing the number of notes getting out there. And we think that AI could be a potentially good way to do this. Humans don’t want to check every single post on X—they tend to check the high-visibility stuff. But machines could potentially write notes on far more content,” said Keith Coleman, X’s VP of product and head of Community Notes, in a statement to ADWEEK.
These fact-checking AI chatbots can be created using X’s Grok AI or OpenAI’s ChatGPT via an API. That said, given AI chatbots’ tendency to hallucinate, these AI-generated community notes will be vetted by humans before broadly appearing on X. AI Community Notes will also be subject to X’s scoring mechanism to reduce inaccuracies.
Even Community Notes leaders agree on the use of AI, as mentioned in X’s papers published on Monday. The paper discusses how AI can help scale Community Notes on the platform. With humans in the loop, it can retain the same trust among users as it currently holds.
What Could Go Wrong with AI-Written Community Notes
X’s implementation of Community Notes is a popular method for fact-checking, to the extent that even Meta opted for this system over its third-party fact-checkers. But the inclusion of AI does raise a lot of eyebrows. Not just because it is prone to making up new realities, but that it can be manipulated as well.
X’s own Grok went on a frenzy, talking about white genocide in every interaction out of nowhere. And there is no saying that the same can’t happen with Community Notes. In a world increasingly embracing AI slop, I am personally not a fan of this move. But who knows, maybe AI-written Community Notes might curb the spread of misinformation on X.
The pilot program has already started, but it will be a while before you start seeing AI-written community notes on X. It also raises concerns about yet another area where humans could be replaced by their AI counterparts. But Coleman has assured that this is not the case. Both humans and AI systems will be crucial for Community Notes.
