AI Generated Image
Meta Pauses an AI Training Program After Data Leak Sparks Employee Ire
Business Insider • • business
Key Points:
- Meta has paused its internal AI training program, the Model Capability Initiative (MCI), after a data leak exposed employees' private conversations, performance data, and transcriptions accessible across the company.
- The incident was rated SEV 2 on Meta's severity scale, where 0 is the most severe, and the company is currently investigating the breach with no indication of malicious access so far.
- MCI, launched in April to improve AI models using employees' keystrokes and mouse movements, faced internal backlash due to privacy concerns before the leak occurred.
- Employees expressed frustration over the data not being sufficiently secured from the start, highlighting ongoing trust issues within Meta regarding data privacy.
- This leak adds to recent security challenges for Meta, including a chatbot vulnerability that compromised Instagram accounts and a rogue AI incident earlier this year.
