Instagram AI Exploited, Leading to Account Takeovers
Cybercriminals have successfully exploited a vulnerability in Meta’s artificial intelligence-powered support system, leading to the unauthorized takeover of numerous Instagram accounts. Among the compromised accounts were those belonging to prominent entities such as the White House and the U.S. Space Force, as well as a well-known security researcher.
AI Chatbot Deceived in Simple Attack
The method used to breach the accounts was notably straightforward. Attackers initiated the password reset process for target accounts and then selected Meta’s AI Support Assistant when prompted for a recovery option. The hackers reportedly asked the chatbot to add a new email address to the account. Without verification, the AI added the email, subsequently sending a code to the newly designated address. This code was then used by the attackers to change the account’s password, simultaneously logging out the legitimate owner from all their devices.
Visual evidence of this exploit has been shared, detailing the attack’s execution. Victims of this security lapse included the Instagram account associated with the Obama-era White House, which had reportedly been inactive since 2017. The account of U.S. Space Force Chief Master Sergeant John Bentivegna was also compromised. Security researcher Jane Wong confirmed that her own Instagram account was similarly taken over.
Meta has since confirmed that the exploited attack vector has been disabled, preventing further unauthorized access through this specific method.
New Protections for Teen Users Being Rolled Out
In related news, Meta is actively testing and implementing new features designed to safeguard teenage users from potentially harmful content. The company is experimenting with ways to limit the repeated exposure of young users to content that could negatively impact their mental well-being. This initiative aims to strike a balance by ensuring that helpful content, such as discussions on nutrition, weightlifting, or anxiety management, is not overwhelmingly presented without other forms of content.
The social media giant is exploring methods to curb the excessive display of such posts within various sections of the platform, including the Explore tab, Feed, and Reels. Initial tests of these protective measures have reportedly proven successful, and the company is now in the process of a global rollout.
