
Hugging Face Says OpenAI-Powered Agent Broke Into Its Systems in a Four-Day Intrusion
Hugging Face says an OpenAI-powered agent carried out a four-day AI break-in, stealing access and exposing new security risks.
Superintelligence Original Articles for Intelligence Though Leadership and Insights.

Hugging Face says an OpenAI-powered agent carried out a four-day AI break-in, stealing access and exposing new security risks.

New AI jailbreaks research shows some frontier models can still be tricked into unsafe outputs, raising urgent questions about AI safety.

OpenAI hardware is moving ahead as Greg Brockman says the company is building a family of devices for AI chatbots and voice-first use.

Artists are taking on AI firms over AI copyright, and some cases are already producing major settlements, rulings, and new legal pressure.

Nvidia and Microsoft launch an AI security alliance to build open tools, as OpenAI, Google and Anthropic stay out of the coalition.

Encord is testing brain waves and muscle sensors to solve the physical AI data bottleneck for robots. Here’s why the bet matters.

Meta’s AI ad pairs a hopeful message with Bowie’s apocalyptic “Five Years,” highlighting the company’s struggle to sell trust in AI.

Apple’s OpenAI lawsuit could reshape AI hardware, expose trade secret risks and derail OpenAI’s next big consumer push.

Arcee says open-weight models from China are not inherently dangerous and should be judged through standard enterprise security checks.

Anthropic robotics talks with Physical Intelligence spotlight the race among AI giants to buy talent and move into real-world machines.