The security developments getting readers’ attention this week, including AI safety, attack automation and new CISO appointments.

During the incident, an early version of Claude Opus 4.6 reached the open internet after a configuration error connected it to real systems, despite being told it was operating in a simulation without internet access.

Google Threat Intelligence Group says adversaries are moving from using models for isolated tasks to deploying agents that can scan targets, respond to errors, and continue operating with limited human input.

Marriott International, Allied Universal and the State of California are among the organizations welcoming new Chief Information Security Officers this month. Find out more about their backgrounds and what they’ll focus on next.

Google says China-linked hackers are moving from basic AI prompting to agents that automate large portions of cyber intrusions. Some campaigns can now be conducted in less than six hours, reducing the time human operators need to spend actively hacking.