MetaCity's 'Report Abuse' Button Has Been Filing the Report Against the User Who Pressed It — Not the User Being Reported — Due to a Reversed Argument in the Moderation API Deployed 11 Days Ago

A reversed source and target argument in MetaCity's moderation API endpoint has caused every 'Report Abuse' submission made through the standard platform interface for the past 11 days to register the reporting user as the subject of the complaint, rather than the user being reported. An estimated 340,000 abuse reports were filed during this period. All 340,000 have been logged against the accounts that submitted them. Approximately 12,000 of those accounts have been automatically flagged by MetaCity's abuse-threshold detection system as a result of the incorrectly attributed reports. Forty-one accounts were automatically suspended. All 41 suspended accounts belonged to users who had been actively reporting harassment and policy violations against themselves.

The 41 suspended accounts are the sharpest illustration of what this bug produced. Each of those users had been reporting harassment or policy violations against themselves — exactly the behavior MetaCity's moderation system is designed to encourage. They were engaging with the platform's safety tools in good faith, repeatedly, because their situation warranted it. The API reversal meant that every report they submitted was recorded as evidence of their own misconduct. The more diligently they reported, the more abuse reports accumulated against their accounts. The automatic suspension threshold is designed to catch serial bad actors. The 11-day bug period turned it into a mechanism for suspending the people who most actively tried to use MetaCity's safety infrastructure. All 41 have now been suspended from the platform while whatever was being done to them continued without any reports logged against the perpetrators.

The clean moderation queue is the other dimension of this that hasn't been fully addressed in MetaCity's statement. 340,000 reports were filed. None of them reached the accounts they were intended for. That means 340,000 potential policy violations — harassment, threats, rule-breaking content, whatever individual users judged worth reporting — went unreviewed during an 11-day window. The moderation team saw a quiet period. The actual moderation backlog, properly attributed, is enormous. MetaCity has said it will correct the report records. It has not addressed what happens to the 340,000 reports that were misdirected — whether those will be retroactively reviewed, whether any of the reported accounts will face any review for the conduct that generated the original complaints, or whether the 11-day window is effectively an amnesty period for policy violations that happened to be reported during it.

You Filed the Report. The System Logged It Against You. You Were Then Suspended for Filing Too Many Reports.

The infrastructure update context matters here. The moderation API reversal wasn't a random corruption — it was introduced by a deliberate code change as part of a broader update. That update passed whatever review and testing process MetaCity applies to moderation system changes before deployment. A test of the report submission flow — submitting a report and verifying which account it was logged against — would have caught this immediately. The fact that it ran for 11 days undetected suggests either that test wasn't part of the deployment review process, or it was performed in an environment that didn't accurately reflect production behavior. MetaCity has corrected the bug. It has not described what review gap allowed a reversed-argument moderation API to reach production, operate for 11 days, and result in 41 users being suspended for reporting abuse against themselves.