On 17 October 2026, Meta suspended the Instagram account of the No Kings Coalition at approximately 1:30pm Eastern Time, during a nationwide campaign coordinating almost 3,000 events aimed at encouraging early voting in the 3 November midterm elections. The organisation lost its primary communication channel with participants at the moment it needed it most. Meta's Vice President of Communications, Andy Stone, initially blamed the suspension on the organisation's own posting, sharing, and commenting activity, claiming high-frequency behaviour had triggered automated enforcement systems. The No Kings organisers disputed this, saying they had published only a handful of posts and were livestreaming events. Given that Meta's own activity logs could resolve this question in minutes, the company abandoned the explanation within 24 hours. Stone's revised statement shifted the cause entirely — from the account's publishing behaviour to its reading behaviour. Meta now claimed its network security systems had detected "the use of a Claude tool reading profiles, posts, and hashtags" at approximately four times the normal rate in "intense bursts." The 4x figure was presented without baseline numbers, measurement period, or threshold context. Whether 4x means 400 requests or 400,000 requests remains unstated. Meta named Anthropic's Claude AI but did not identify the specific tool, its operator, or whether the No Kings organisers had authorised it. The account became publicly visible again after approximately 90 minutes, but the organisers remained locked out until 5:33pm Eastern Time — roughly four hours after the suspension. Stone described the account as "restored" after the initial 90-minute window, language that obscured the fact that the people managing it still could not publish posts or communicate with supporters. Meta's own automated review had already determined the suspension was wrong, yet administrative access required additional human authorisation that Meta has not explained. The author of the analysis — a cybersecurity founder who built the first security chatbots between 2015 and 2019 — makes a structural point: automated requests are not inherently malicious. Search engines, monitoring tools, research applications and AI assistants make billions of automated requests hourly. Platforms have rate-limiting and credential-restriction systems designed to handle excess traffic without suspending accounts. Meta has not explained why it chose full account suspension over simply throttling the automated requests, nor whether it applies the same restrictions when its own AI assistant, Muse, automates interactions on Instagram. The deeper issue is accountability architecture. The No Kings Coalition is one of America's most prominent anti-Trump movements. Suspending its account during a major mobilisation was a politically significant action with predictable consequences. The author argues that given the organisation's prominence and the public backlash, multiple Meta teams and senior managers would have been involved in decisions about both the suspension and the delayed restoration. Meta's framing of the incident as a software error obscures the human decisions embedded in the process — who reviewed it, who authorised restoration, and why organisers were kept locked out after the system had already flagged the suspension as incorrect. Meta admitted the suspension was wrong. What it has not done is provide the actual request counts, name the individuals who made decisions, explain why full suspension was chosen over rate-limiting, or clarify why administrative access was withheld for hours after the error was identified. The company whose ad-targeting systems can map a child's preferences from a relative's pet photos says it cannot explain its own enforcement actions against a political organisation during a national election campaign.
