WhatsApp’s effort to stop bad actors briefly swept up legitimate users, triggering unexplained account suspensions that the company now admits were mistakes.
WhatsApp has acknowledged that some users were incorrectly suspended after their activity was flagged as violating the company’s terms. While the reviews typically last around 24 hours, affected users temporarily lost access to messaging, calls, and their conversations.
The incident underscores a growing challenge for digital platforms that rely on automated enforcement: balancing detection with the risk of false positives.
A WhatsApp bug hit legitimate users hard
According to The Independent, over the past few days, several WhatsApp users have taken to social media to complain that they could no longer access their WhatsApp accounts.
Affected users said the only message they saw when they opened WhatsApp was: “Your account activity and device info is being checked to make sure it follows our Terms of Service. We’ll notify you of the result typically within 24 hours.”
Anyone who’s had their WhatsApp account suspended would be familiar with the message above, which suggests they’ve violated WhatsApp’s rules. However, in this case, the users say they did nothing wrong.
According to The Independent, WhatsApp allows users to appeal a ban once, and after that, it becomes permanent. Although the messaging platform says that its moderation systems erroneously enforce bans, it argues that that ability is in place to protect WhatsApp users from harm.
The company also says it had restored accounts that were accidentally flagged.
Not the first time
Last August, Meta faced a similar backlash after some Facebook and Instagram users said their accounts were suspended over alleged child sexual exploitation, abuse, or nudity violations despite insisting they had done nothing wrong.
Meta, which increasingly relies on AI to help moderate content across its platforms, says its enforcement process also includes human reviewers.
Meta is not the only company grappling with the limits of automated moderation. Last month, Discord acknowledged a bug in its automated enforcement systems after several users were mistakenly banned for posting benign content, prompting the platform to reverse the erroneous actions.
Where does this leave everyone?
Taken together, these incidents point to a broader shift in how online platforms are managed.
As platforms record millions and billions of users, companies are increasingly turning to AI and automated decision-making to enforce policies, detect abuse and fraud, and respond faster than human teams ever could.
That approach is becoming less of a competitive advantage and more of an operational necessity. Still, it also means software bugs, model errors, or overly aggressive detection thresholds can affect large numbers of legitimate users almost instantly.
Such wrongful suspensions can disrupt businesses, delay critical communications, or temporarily cut users off from services they depend on, often with little explanation beyond an automated notification.
For enterprises and software developers, automated systems need transparent review processes, rapid appeal mechanisms, and continuous monitoring for false positives.
Also read: Meta is removing Instagram videos made with Ray-Ban smart glasses when they are used to harass or exploit strangers, with repeat offenders facing potential account bans.