When AI agents look like attackers: what behavioral telemetry tells us

Sophos X-Ops blog roundup highlighting how AI is changing attack dynamics and detection: AI coding agents can behave like human adversaries and trigger endpoint telemetry, and AI can rapidly turn patched bugs into working exploits—underscoring gaps in exploit mitigation. Other posts cover a ransomware partnership (Vect and TeamPCP) leveraging credentials from supply-chain compromises, an AI threat taxonomy, use of AI to hunt infostealers, and a large June Patch Tuesday (>500 CVEs). Operational takeaways include strengthening endpoint telemetry and default-on exploit mitigations, urgent patch/v