What you need to know
- Meta’s AI reportedly aided hackers in hijacking users’ Instagram accounts, and the method was embarrassingly simple.
- Hackers took screenshots, showing that need only ask the AI for access to an account, provide their email, and request a reset passcode.
- Many users were affected by this; however, a spokesperson claims this issue has been fixed.
There’s been a harrowing discovery about Meta’s AI that is not only dangerous, but it’s also causing an uproar amongst users over account security.
This most likely isn’t the way Meta wanted to start June. A video report from 404Media says that anyone can ask Meta AI for access to an account that’s not theirs, and it’ll do it. The publication says that the AI is offering access to “high-profile” Instagram accounts when asked. This has reportedly become a trend with hackers over the past few months: talking with the AI to gain access to accounts that aren’t theirs.
The method was embarrassingly simple. The video report’s screenshots show that a hacker could tell Meta AI that “they want access to a specific account.” What’s more, they would offer their email address, then ask for a password reset code, and the AI would willingly provide that crucial security code.
It was reiterated that Meta changed how it works its customer support. Users aren’t speaking to another human. Instead, back in March, the publication states Meta went from human workers to AI for customer support—and this is the result. With further research, 404Media says it’s encountered several users whose accounts have been hacked using this easy method.
Moreover, several high-profile accounts, such as former President Barack Obama’s White House account, the clothing brand Sephora, and others, were supposedly seized by hackers after enlisting Meta AI for help.
An embarrassing problem
BBC got involved with its own report on the matter. It reports that a Meta spokesperson, Andy Stone, posted on X that the “issue has been resolved and we are securing impacted accounts.” However, Stone also threw shade at allegations that Meta’s AI was responsible for helping hackers hijack accounts from world leaders. It seems that this is being positioned as a separate issue entirely.
Users across social media are voicing their displeasure with Meta after this incident. Many question why they would resort to using AI for such a crucial area of support and not have more appropriate verification measures in place. It comes as no surprise that Meta is so involved with AI. The company’s Q1 2026 earnings report showed that its spending increased by 35% ($33 billion), which was all for AI. More importantly, Meta had another massive wave of layoffs, letting go of 8,000 employees in May.
Those layoffs seemed to have the echo of AI looming in the background.
Android Central’s Take
Appalling is the one word that comes to mind for this. If a company is going to invest so much money into AI and let go of thousands of real humans just for the sake of this tech, shouldn’t it take its decision more seriously? Sure, the issue has been fixed. It still happened in the first place when it shouldn’t have. Who knows why Meta is so fixated on AI (and not humans). Whatever the case may be, the trust is gone.