AI is already writing more than half of X’s helpful Community Notes, and that changes everything about who gets fact-checked, what gets fact-checked, and what happens when “the community” starts looking a lot more synthetic. Alexios Mantzarlis, co-founder of Indicator and a self-described “recovering fact checker,” breaks down the surprising rise of AI-written notes, why bots sometimes outperform humans, and the bigger danger hiding underneath the numbers: a moderation system that can look healthier on paper while quietly losing human participation, language diversity, and political coverage.  Hosts Jim Ludes and Mark Jacobson also unpack the recent migrant crisis in Ceuta, Spain, and increasing evidence of adversarial manipulation.

<p>AI is already writing more than half of X&rsquo;s helpful Community Notes, and that changes everything about who gets fact-checked, what gets fact-checked, and what happens when &ldquo;the community&rdquo; starts looking a lot more synthetic. Alexios Mantzarlis, co-founder of Indicator and a self-described &ldquo;recovering fact checker,&rdquo; breaks down the surprising rise of AI-written notes, why bots sometimes outperform humans, and the bigger danger hiding underneath the numbers: a moderation system that can look healthier on paper while quietly losing human participation, language diversity, and political coverage.&nbsp; Hosts Jim Ludes and Mark Jacobson also unpack the recent migrant crisis in Ceuta, Spain, and increasing evidence of adversarial manipulation.</p>

From the Pell Center at Salve Regina University and the creators of the Active Measures Newsletter, a weekly dive into the latest trends in political warfare, influence, and information campaigns. 

<p>From the Pell Center at Salve Regina University and the creators of the Active Measures Newsletter, a weekly dive into the latest trends in political warfare, influence, and information campaigns. </p>

A weekly dive into the headlines in the Active Measures Newsletter from the Pell Center at Salve Regina University.

The Active Measures Newsletter Podcast

AI bots are now writing more than half of X’s helpful community notes, and that shift may be quietly reshaping how truth is judged online. In this episode, Alexios Mantzarlis explains why machine-written fact checks can outperform humans in some cases, where they fail on breaking news and politics, and how platform design may be pushing AI contributors toward safer, less consequential targets. The conversation also explores who is building these bots, why Grok matters, and how tactics like “no notes needed” can be used to game the system. Together, the hosts unpack the growing tension between crowdsourced truth and the risk of automating it into something easier to manipulate.