Meta's automated protections wrongly suspended No Kings coalition's Instagram account, blamed on Claude tool traffic
Meta said its automated security systems incorrectly took the No Kings coalition's Instagram account offline on October 17. Meta blamed sharp bursts of reading activity by an unspecified Claude tool that resembled bot traffic. The account was suspended for about 90 minutes while nearly 3,000 events were taking place across the US, and administrators could not access the backend for about four hours. Meta had first attributed the suspension to high-frequency posting before changing its explanation, and the coalition demanded Meta answer for the suspension.
What the AI did
Meta said its network security systems detected an unspecified Claude tool reading Instagram profiles, posts and hashtags in intense bursts on October 17, with the reading volume rising about fourfold. According to Meta, this activity looked like automated bot traffic and caused its automated protections to wrongly take the No Kings coalition's Instagram account offline. Meta has not said which Claude tool was involved, who was operating it, or what it was doing in connection with the account.
First reported October 18, 2026 · Added to the register October 31, 2026 · 1 source
- Developer
- Anthropic
- Deployer
- Meta
- Model
- Claude tool
- Affected
- No Kings coalition
- Country
- United States
- When it happened
- October 17, 2026
- First reported
- October 18, 2026
What this means for you
Could this affect you?
If you rely on Instagram or other Meta platforms, especially with AI tools connected to your accounts. Any organisation that depends on Instagram or other Meta platforms for time-critical communication, particularly one using AI agents or outside tools to read or manage its accounts, could be cut off by the same kind of automated enforcement that suspended the No Kings coalition during nearly 3,000 events.
What to check
- Check whether AI tools connected to your social media accounts generate high-volume automated traffic.
- Keep backup communication channels ready for critical campaigns.
- Document escalation routes with platforms for challenging wrongful automated suspensions.
Areas of your AI programme this touches
Timeline
- October 17, 2026Happened
- October 18, 2026First reported
Every fact and its source (5)
- Date of activity bursts and suspensionOctober 17“in intense bursts on October 17”[1]
- Suspension durationAbout 90 minutes“The account was suspended for about 90 minutes while nearly 3,000 events were taking place across the US”[1]
- Loss of admin backend accessAbout four hours“could not access the backend for about four hours”[1]
- Increase in Claude tool reading volumeAbout 4x“the rate of which sharply increased (~4x volume)”[1]
- RestorationAutomated re-review restored the account an hour later“An automated re-review of the removal action an hour later recognized”[1]
Sources
- Meta blames Claude AI tool for wrongly suspending US protest group's Instagram accountmedianama.com · October 19, 2026
How this record is classified. Severity S (3/5): real harm to one organisation, data exposed, or a containment failure. OECD level: incident, an event in which an AI system led to actual harm. Evidence: Confirmed, meaning a company, official or court statement.
The facts in At a glance and in Every fact and its source are quoted from the sources listed. The summary, What the AI did and What this means for you are written from those sources and checked against them automatically before publication. Records reflect what has been disclosed, not everything that has happened. Incidents are found through news feeds and the GDELT Project.
