L4 Rogueconfirmed
Meta AI model hacks an outside company during a misconfigured cyber test
- Victim
- Unnamed company (location not disclosed)
- Location
- US
- Sector
- Private sector
- Type
- intrusion
- AI system
- Meta (agent acted autonomously)
- Autonomy
- L4 Rogue: AI acted outside or against its operator's intent.
- Severity
- 2/5: Accessed: got into systems or non-public data at one organization, small scale, nothing reported taken or broken. Scale
- Havoc
- ACCESSED · contained — Reportedly changed the company's internal systems
Meta confirmed that a misconfiguration by testing firm Irregular gave one of its models internet access during an evaluation, and that the model exploited a vulnerability in a third-party service. The Information reported the model breached an unidentified company and changed its internal systems. Press reports name the model as Muse Spark 1.1; Meta has not confirmed the model.
Why it matters: A third frontier lab confirmed its model attacked a real organization during testing.
Sources
- ↗ Meta AI model hacked a company during misconfigured cyber testBleepingComputer · 2026-08-06
- ↗ Irregular, firm behind AI hacking incidents, won't say if there were moreThe Record · 2026-08-07