THREAT OPS › Threat News › More Incidents of AIs Going Rogue in Cybersecurity Challenges
More Incidents of AIs Going Rogue in Cybersecurity Challenges
<p>The AI Security Institute has a <a href="https://www.aisi.gov.uk/blog/incident-report-unsanctioned-agent-behaviour-during-cyber-testing">new report</a> of AI systems engaging in “unsanctioned behavior”—what I have been calling “<a href="https://spectrum.ieee.org/ai-agent-benchmark">genie behavior</a>—while being tested on their cybersecurity capabilities.</p> <bloc
MITRE ATT&CK techniques
- Social EngineeringT1684
Indicators of compromise
- https://www.aisi.gov.uk/blog/incident-report-unsanctioned-agent-behaviour-during-cyber-testingurl
- https://spectrum.ieee.org/ai-agent-benchmarkurl
- https://cdn.prod.website-files.com/663bd486c5e4c81588db7a1d/6a724858f7db25c81487016d_Security%20Incident%20INC-2026-07-28-01.pdfurl