Hackers Exploit Meta AI Support Tool to Take Over Obama, Sephora and Space Force Instagram Accounts

Several high-profile Instagram accounts were compromised over the weekend after hackers exploited a vulnerability in Meta's AI-powered customer support assistant, raising fresh concerns about the growing role of artificial intelligence in account security. According to multiple reports, including investigations by The Guardian, 404 Media and cybersecurity researchers, pro-Iranian hacking groups leveraged the flaw to seize control of prominent accounts linked to the Obama-era White House, Sephora and the United States Space Force. Rather than breaching Meta's internal systems directly, attackers reportedly manipulated the company's automated recovery chatbot into changing account recovery information. The incident has reignited debate over whether AI systems should be entrusted with sensitive security functions that were once handled by trained human support personnel.

Investigators say the attack relied on a surprisingly simple form of social engineering. After masking their locations with virtual private networks designed to imitate the geographic region of the targeted account holders, hackers initiated conversations with Meta's AI support assistant through the account recovery process. The chatbot allegedly allowed attackers to replace the email address associated with a victim's Instagram profile without requiring sufficient identity verification. Once the recovery email had been changed, hackers simply requested a password reset and received the reset link directly in their own inboxes, effectively locking out the legitimate account owners. Cybersecurity researchers described the flaw as an example of a so-called prompt injection attack, where users manipulate an AI system into performing actions beyond what its security safeguards were intended to allow.

Meta acknowledged the vulnerability after videos demonstrating the exploit began circulating on Telegram and X. The company said it had patched the flaw and was working to restore affected accounts. Responding publicly to reports of the incident, Meta spokesperson Andy Stone wrote: «This issue has been resolved and we are securing impacted accounts.» The company has not disclosed how many accounts were affected before the vulnerability was fixed. The breach is particularly embarrassing for Meta because the AI support assistant was launched earlier this year as part of a major effort to modernize account recovery and customer support across Facebook and Instagram. The company had promoted the technology as a faster and simpler way for users to recover access to their accounts and resolve security-related issues without waiting for human assistance.

Among the most prominent victims was the archived Instagram account associated with Barack Obama's White House. According to reports cited by TMZ and cybersecurity researchers, users noticed unusual activity on the account after it began publishing pro-Iranian content and political propaganda. One of the most widely circulated posts reportedly featured an AI-generated image accompanied by a caption translated as: «The White House is under Shiites' control.» Screenshots of the content quickly spread across social media before Meta removed the posts and restored access to the account. Researchers noted that the compromise appeared intended more as a political statement than a financial operation, although stolen Instagram accounts are frequently sold on underground markets for significant sums. The incident demonstrated how a vulnerability in an automated support system could quickly become a geopolitical messaging tool in the hands of hostile actors.

The official Instagram account of Chief Master Sergeant of the Space Force John Bentivegna was also compromised during the same wave of attacks. Hackers used the account to publish a series of pro-Iranian Instagram Stories and propaganda images before the content was removed. The material reportedly included anti-American messaging, references to geopolitical tensions in the Middle East and audio associated with psychological warfare campaigns. Confirming the breach, a Space Force spokesperson stated: «The official Instagram account of the Chief Master Sergeant of the United States Space Force has been compromised. We are currently working with the appropriate teams to regain access and resolve the issue as quickly as possible.» The military branch declined to provide further details regarding the nature of the attack or whether additional government-affiliated social media accounts had been targeted through the same vulnerability.

Sephora was among the corporate victims caught in the campaign. Hackers briefly gained control of the official Sephora Collection Instagram account and published explicit graphic content during the early hours of June 1 before the posts were deleted. Shortly after regaining access, the company posted a temporary Instagram Story informing followers that the account had been secured. The incident highlights the broader risks facing organizations that rely heavily on social media platforms for customer engagement, branding and marketing. Security experts warn that one of the first major examples of how AI-driven customer service can be weaponized against the very users it was designed to protect.

Created by humans, assisted by AI.