News

Meta fixed a security flaw in Meta AI that exposed private prompts. No abuse was found. Researcher earned $10,000 reward.