Taking the temperature of AI.
12 stories
Enterprise customers appear willing to switch between frontier AI labs as new models land, raising questions about how durable their spending commitments really are.
Multiple cybersecurity researchers said they suddenly lost access to OpenAI’s Trusted Access for Cyber program, which offers vetted users models with fewer guardrails.
The new offering adds age-appropriate safety measures, parental controls, and learning tools to the chatbot teens were already using.
The new dedicated mode centralizes existing youth safeguards and adds fresh safety features as regulators and rival platforms tighten age-based protections for AI tools.
Anthropic and OpenAI publish snapshots of how people use Claude and ChatGPT, but outside researchers say the public only sees company-selected data.
A watershed cybersecurity breach has forced the AI lab to confront hard questions about the culture that enabled it.
The AI company is tapping Dali Rajic to become its top salesperson as leadership changes continue.
On TechCrunch's Equity podcast, Altman's plea to slow the rate of AI progress puts him at the center of the industry's acceleration-versus-deceleration fight.
The AI lab is reportedly building a system aimed at sustaining work beyond the single prompt.
A report says the incident helps explain how OpenAI hacked into Hugging Face, with 10 days passing between exploitation of the JFrog flaw and a released fix.
The initiative is aimed at accelerating scientific research, collaboration and discovery.
Retaining reasoning and enabling compaction boosted GPT-5.6 performance and efficiency on the abstract-reasoning benchmark.
