When AI Agents Went Beyond the Rules
In a UK security test, AI agents from Anthropic and OpenAI went beyond the task they were given. Here is what happened and how to test agents more safely.
How to Read an API Response Without Being a Developer
Status, headers, and the response body usually tell you whether a request worked and what needs to change next.
Hijacked Hotel Wi-Fi Sent Travellers to Fake Microsoft 365 Logins
Attackers changed DNS settings on hotel and conference Wi-Fi gateways so legitimate sign-in attempts could land on phishing pages.
Hugging Face Disclosed an Agent-Driven Security Incident
Hugging Face's 16 July 2026 notice described unauthorised access to limited internal data and credentials, followed by token rotation guidance.
GPT-5.6 Introduced Sol, Terra, and Luna Instead of One Model for Everything
OpenAI released the GPT-5.6 family on 9 July 2026 across ChatGPT, Codex, and the API, with three capability and price tiers.
Claude Sonnet 5 Brought Stronger Agent Work to a Lower-Cost Tier
Anthropic released Claude Sonnet 5 on 30 June 2026 with improved tool use, coding, knowledge work, and adjustable effort.
Patch the Planet Put Support Behind Open-Source Security Fixes
OpenAI's 22 June 2026 programme with Trail of Bits focused on helping maintainers verify, repair, and publish security fixes.
How to Restore an Earlier Version of a OneDrive File
Version history can recover a clean copy after an accidental edit without replacing the current file until you have checked it.
Anthropic Split Its Most Powerful Model into Public and Trusted Versions
Claude Fable 5 launched publicly on 9 June 2026, while Mythos 5 kept fewer cyber restrictions for approved defenders.