Meta spokesperson Andy Stone said on X that the account had used a Claude AI tool to read profiles, posts, and hashtags in intense bursts that looked like automated bot traffic.
He said automated review caught the error within an hour, but additional hurdles kept admins locked out.
No Kings says it used Claude as a keyword and hashtag search tool to track conversations across social media.
A coalition of 14 state attorneys general asked Meta to explain the suspension and detail how it will keep legitimate election activity from being disrupted, citing AI agents marketed to consumers as compatible with social media accounts.
Sen. Chris Van Hollen and Reps.
Robert Garcia and Jamie Raskin also pressed Meta for answers.
No Kings rejected Meta's offer of a meeting with a technical expert, and after meeting Meta executives on October 30, said the company has "no plans" to change policies that cause harmful over-enforcement.
Ars reported it found no evidence that the Trump administration asked Meta to restrict the account or that anyone at Meta deliberately targeted it.
Skeptics point to Meta's record on election content and its closer ties to Trump. Ezra Levin of Indivisible, a coalition member, called the technical explanation hard to accept and wants Meta to disclose any communications with the administration. The Electronic Frontier Foundation's David Greene called the timing suspicious and the explanation incomplete, and suggested human review of election content and emergency support channels. Kate Ruane of the Center for Democracy and Technology said Meta should explain why its mistake-prevention systems did not trigger. Meta said it would publish a public reply, but it did not respond to Ars' request for comment.