Discord unbans 8,000 users after AI wrongly flagged harmless images

Discord has reinstated more than 8,000 accounts it accidentally banned. A fault in its automated moderation system incorrectly flagged harmless uploads as illegal content. It stumbled over spreadsheets, chessboards, and game textures. The bans were permanent and without any human approval.

In a thread posted by its support account on X on July 7, 2026, Discord detailed the errors that occurred.

Users had been encountering red flags since May. Discord’s team was able to identify the issue and distribute a fix after ~200 additional accounts were detected over the weekend. The company has announced that all affected accounts are currently being reinstated.

Similarity matching flagged spreadsheets and grids

Every uploaded file is compared to databases of content that have already been identified as dangerous by Discord’s security tools. The company acknowledges that similarity matching, the basis for that comparison, has the potential to misfire and produce false positives.

On paper, before an account is penalized, a member of the Trust and Safety team should review anything the system flags.

That step was eliminated by the bug. The system completely banned accounts rather than sending dubious uploads to a reviewer. Simple photos caused the filter to trip, resulting in instant suspensions. Along with the spreadsheets and chessboards, transparent white and gray backgrounds were also captured.

Before Discord provided an explanation, the annoyance appeared on X and Reddit. Posting pictures with square grid patterns was the only reason why people reported losing their accounts.

A theory was put forth by several users. The detection model may have become agitated around anything that resembled a grid because grid-like layouts have previously been used to conceal NSFW and child-exploitation content from automated scanners.

“…We should have caught this sooner. We’re working on better safeguards so this can’t happen again,” Discord wrote in its X thread.

This weekend, our safety systems incorrectly triggered and banned around 200 accounts. Everyone affected has been reinstated.

Here’s what happened. A thread 🧵 https://t.co/wXH32iiv8I

— Discord Support (@discord_support) July 7, 2026

The same bans hit Meta and Tumblr

On July 4, a game director going by the handle JDBRYANT claimed that his account was deleted because the automod interpreted his game textures as depicting child sexual abuse.

A channel he uses for all of his work-related communications was cut off by the ban. He claimed to have submitted a request for the suspension to be reviewed.

Another user went further. “…This vulnerability needs to be fixed immediately with all false bans overturned. Losing a Discord account to something as unfair as this can be extremely devastating and affect users severely, and every day millions of users are affected by false AI bans. This needs to be stopped,” the person wrote on X.

Discord has over 200 million unique users every month, according to the company’s own statistics that Cryptopolitan cited in February.

Instagram and Facebook Groups users reported waves of unexplained suspensions last year as well, widely blamed on automated moderation, though Meta never confirmed whether AI errors were behind them. Meta’s Oversight Board has since pressed the company for more transparency.

Tumblr fielded similar complaints over mass suspensions with no clear reason.

Discord spent early 2026 rolling out global age-verification and teen-by-default settings, framing the changes around child protection, as Cryptopolitan reported in February.