The Meta platform has discovered that the AI agent, Agjenti AI, is causing massive leaks of sensitive employee data. This alarming situation has prompted immediate action to contain the breach and investigate its origins. The extent of the compromised information is currently being assessed, but initial reports indicate a significant volume of confidential data was exposed. Meta is taking this matter extremely seriously and is working diligently to secure its systems and prevent further data loss.

Meta’s artificial intelligence agent caused a significant breach, exposing sensitive user data to AutoTech personnel. The Express Gazeta detailed the incident on March 20, 2026, at 22:50. An AI agent developed by meta provided an engineer with instructions that ultimately led to the disclosure of considerable confidential user and company data.

This exposure lasted for roughly two hours, involving several other staff members. Meta has confirmed the event transpired when an employee sought assistance through an internal forum concerning an engineering problem. The AI agent offered a proposed solution, which the employee subsequently executed, triggering a temporary release of vital data.

A meta spokesperson clarified that “no user data was compromised,” stressing that inaccurate advice could have been provided by anyone. The situation generated considerable security concern within the organization. The incident highlighted vulnerabilities in the use of AI tools and prompted a thorough review of internal processes.

The focus remains on ensuring data protection and preventing similar occurrences in the future.

Topics: #meta #data #agent