Meta has linked the temporary loss of the No Kings coalition’s Instagram presence to a Claude AI tool. According to MediaNama, Meta spokesperson Andy Stone posted on X on October 19 that the group’s network security systems flagged a sharp rise in "reading" behavior. This included profiles, posts, and hashtags, with volume spiking roughly four times in intense bursts on October 17. Stone noted that these abrupt, high-volume requests resembled automated bot traffic rather than ordinary human browsing, causing the automated protections to take the account offline.
The outage occurred during the coalition’s nationwide push to encourage early voting for the November midterms. Newsweek reported that the account was offline for about 90 minutes while nearly 3,000 events happened across the US. Although an automated re-review restored the public feed an hour later, the organizers faced continued disruption. MediaNama noted that it took additional time to fix the backend, leaving administrators unable to access the dashboard for roughly four hours. A No Kings representative told Newsweek that the delay meant Meta was "actively depriving us of our platform" during a critical window.
Meta’s account of the incident has shifted since the initial outage. On the day of the events, Stone suggested that high-frequency posting or commenting had triggered the automated systems. The later update, however, pointed to the account’s reading activity as the primary cause. Bloomberg reported that Meta has not yet identified the specific Claude tool, the operator, or the exact activity being performed on the account.
The incident highlights ongoing friction between Meta’s automated enforcement and users, particularly in India. MediaNama highlighted cases like Blink Digital, whose four Business Manager accounts were disabled, and Aakanksha Gupta, whose deactivation led to a Delhi High Court order for disclosure. The Internet Freedom Foundation also noted that some October notices to Indian users cited action "through an automated system" without specifying the legal basis. Meta’s Oversight Board warned in June that the company does not always publicly list all restrictions before disabling accounts, urging caution against appeal processes that simply rerun automated decisions.