AI Security Threats: Inbox Hijacking & OpenCode Data Breach
Two distinct AI-related security events surfaced: a distillation attack breached OpenCode Console impacting 1.4% of users, while Barracuda documented phishing emails now targeting AI email assistants rather than humans directly.
What we know
An account on X RT'd OpenCode's disclosure of a 'large scale distillation attack' affecting 1.4% of Console users, with notifications sent via email. In a separate post, another account on X flagged a Barracuda-documented attack vector where phishing emails are now crafted to manipulate the AI assistant summarizing your inbox—bypassing the human reader entirely. Another account on X connected the dots with a 2022 prediction that AI-driven persuasion and disinformation would make social media 'likely not salvageable,' suggesting that prediction is now reality.
By the Global Eye 24 desk, with AI assistance. Updated as the story develops. Spot an error? Write to corrections@globaleye24.com.