Brain Waves as a New Data Source for Physical AI

According to a report from TechCrunch, researchers are exploring brain wave readings as a potential next frontier for training physical AI models. The piece notes that frontier physical AI systems have already moved well beyond consumer video as a data source, now requiring multiple camera angles and dense annotation to function effectively. Brain wave data is being examined as an additional modality that could help these models better understand and respond to human intent in real-world environments. The full article is available at TechCrunch.

Nvidia and Microsoft Launch Open Secure AI Alliance

Nvidia announced on Monday that it is partnering with Microsoft, SpaceX, IBM, and a number of other technology companies to develop and share open-source AI security tools, according to reporting from The Verge. The initiative, called the Open Secure AI Alliance, is centered on the position that open tools are necessary to effectively defend against threats posed by frontier AI models. Notably, the alliance launched without the participation of OpenAI, Google, or Anthropic. The Verge reports that the effort is a direct response to growing concerns over AI-related security vulnerabilities. More details can be found at The Verge.

OpenAI Research: AI Is Expanding the Scope of Work

OpenAI published new research this week examining how AI tools are changing what employees do on the job. According to the OpenAI blog, the findings show that ChatGPT users are increasingly taking on tasks that span multiple traditional job roles, effectively reshaping the boundaries of individual positions. The research frames this as AI expanding the range of work people engage in, rather than simply automating existing tasks. The full research post is available on the OpenAI blog.

These developments reflect continued momentum across physical AI research, industry-level security collaboration, and the evolving role of AI in professional workflows.