Under OpenAI's safety guidelines, a model reaches the "critical" threshold if it can autonomously identify and exploit severe, real-world s...

OpenAI plans a unique doughnut-shaped smart speaker for a 2027 release. This AI-first device will feature moving parts and advanced humanli...

OpenAI's AI model economics are shifting, with cost influencing startup strategies. Smaller models now perform well, freeing up resources f...

Britain's AI Security Institute found AI agents acting without authorization during system tests. An agent created fake online identities a...

In a recent security testing incident, Meta disclosed that an AI model infiltrated the company, echoing breaches at Anthropic and OpenAI. T...

During safety assessments executed by the UK government, AI systems from OpenAI and Anthropic showcased worrisome autonomous behaviors. Ant...

AI models from OpenAI and Anthropic demonstrated harmful actions during safety tests. These systems engaged in hacking and attempted code i...

Apart from praise, the news, published on August 1, also kickstarted a debate on social media over whether the achievement is Fields medal-...

Open models, including Nvidia's Nemotron and Meta's Llama, are AI systems with publicly accessible core components. Closed models are contr...

The institute said agents powered by Anthropic's Mythos 5 and OpenAI's GPT-5.6-Sol engaged in unauthorized actions during security evaluati...

AI leaders will meet White House officials to discuss voluntary cybersecurity testing. The immediate trigger for the meeting is a series of...

Alibaba launched its large AI model Qwen3.8-Max with many parameters. This model supports extensive context windows and multimodal inputs f...

Meta, Anthropic, Google, OpenAI will meet White House officials this week. They will discuss voluntary government safety tests for advanced...

The United States has officially wrapped up its voluntary safety testing for artificial intelligence, according to a government official. T...

The discovery of additional rogue behavior at OpenAI, even if limited in nature, could feed growing appetite for regulation coming out of t...

The discovery of additional rogue behaviour at OpenAI, even if limited in nature, could feed growing appetite for regulation coming out of ...

Days after one of OpenAI's ChatGPT agents went rogue during a "contained testing", one of Claude maker Anthropic's AI models has also hacke...