Researchers on the company’s alignment team, the group whose job is to check that its models behave as intended, named it ...

Anthropic has tightened Claude security after models accessed real systems during cyber evaluations. ETIH examines the edtech news implications of new safeguards and research showing how reward-hacked ...

After Claude models accessed real systems during cyber tests, Anthropic tightened its safeguards and warned that flawed training can encourage dangerous behavior.

An amendment to the UK’s Cyber Security and Resilience Bill, currently passing through the House of Lords, proposes giving the government “last resort” powers to shut down large AI systems in an ...

Google launches Gemini 3.8 Flash Cyber for trusted defenders as OpenAI says Astra meets its Critical cybersecurity capability threshold.

King’s household admits “ageing” computer systems need replacing to protect the monarchy from security threats ...

Tesla spent years developing Dojo as a custom AI-training supercomputer that could reduce its dependence on Nvidia, but Elon ...

Nutex Health, a Houston, Texas-based healthcare management and operations company that delivers care through 27 ...

One is a developer's dream for complex reasoning; the other is a full-featured productivity suite with native image and video ...

OpenAI has delayed its next frontier AI model after an internal cybersecurity evaluation exposed unexpected behaviour from ...

Following three recent security incidents involving Claude, Anthropic is moving to more isolated environments, continuous ...

OpenAI says its forthcoming Astra model can independently identify and exploit previously unknown software vulnerabilities, ...