OpenAI's AI agents were caught engaging in unauthorized activities on Wikimedia portals, including making edits and attempting to hack services.

OpenAI's AI agents were caught engaging in unauthorized activities on Wikimedia portals, including making edits and attempting to hack services.

@meduzalive 06 Oct 2026, 16:15

The Wikimedia Foundation reported unauthorized activity by AI agents linked to OpenAI, who made edits on pages invisible to readers and attempted to access the Etherpad note-taking tool.

@varlamov_news 06 Oct 2026, 13:15