Discord has acknowledged that a bug in its AI-powered content moderation system mistakenly banned more than 8,000 users over the past two months, after harmless images, including spreadsheets, chessboards, game textures, and simple white or gray transparent backgrounds, were incorrectly flagged as harmful content.
The company confirmed the issue had been affecting accounts since May, with an additional 200 users banned over the weekend before its engineering team identified and fixed the problem. All affected accounts are in the process of being restored.
How the bug worked. Discord’s automated safety system uses similarity matching to compare uploaded images against databases of known harmful material. The system is designed to flag potential violations for manual review by a human member of the Trust & Safety team. However, a bug in the pipeline caused the system to bypass the human review step entirely and issue permanent bans immediately.
The false positives appear to have been triggered by grid-like patterns, chessboards, spreadsheets, game textures, which share structural characteristics with techniques historically used to obfuscate prohibited content from automated detection. White and gray transparent backgrounds also produced hash values that incorrectly matched entries in the database.
In a detailed thread on X, Discord’s support team wrote: “Our systems flag content by matching it against known harmful material. This kind of similarity matching can produce false positives, which is why a member of our Trust & Safety team always reviews flagged content before any action is taken. The intended behavior is to… We’re working on better safeguards so this can’t happen again.”
User impact. Affected users included a game director who lost access to their professional communication channel due to game textures being flagged, and members of crypto and NFT communities who rely on Discord as their primary coordination layer. Users reported permanent suspensions with no effective appeals process during the outage.
The incident joins a growing list of high-profile AI moderation failures. Instagram and Facebook Groups faced widespread unexplained bans in 2025, and Tumblr’s automated filtering similarly flagged benign content as mature. Discord’s case is notable for its transparency: rather than remaining silent on the cause, the company published a public postmortem and acknowledged the root cause.