OpenAI Loses Its Only Ethicist, and Isn’t Replacing Her
Chloé Bakalar was OpenAI’s sole dedicated head of ethics. After eleven months, she’s gone, along with two other safety leaders.
AI & Society
AI News
Gemini’s Growth: Why User Numbers Do Not Crown an AI Winner
Google reports rapid Gemini growth while market studies show shifting shares. Why big AI user numbers make sense only when their measurement method is clear.
Business & Investments
Grok 4.6: Why Price Is Becoming an AI Feature
Grok 4.6 pairs frontier ambitions with low API pricing. For teams, the key question is no longer the model name but cost, speed, and control in daily work.
Chips & Hardware
Nvidia’s Nemotron 3.5 Lightning: Small, Fast, and Open by Design
Nvidia’s Nemotron 3.5 Lightning is a compact open-weight AI model built for speed over raw intelligence, and it reveals a clear business strategy behind it.
AI Research
Researchers Crack the Encrypted Thoughts of ChatGPT, Claude, and Gemini
Security researchers extracted encrypted reasoning traces from ChatGPT, Claude, and Gemini, finding passwords and API keys that were never meant to be seen.
Business & Investments
ChatGPT Business Gets More Expensive: What Premium Seats Change
OpenAI is planning Premium Seats for ChatGPT Business. Up to $125 a month buys mostly more capacity. What teams should take from it.
Regulation & Privacy
Hidden AI Reasoning Traces: Why Encryption Alone Is Not Enough
A new study shows how encrypted reasoning data from AI APIs can become a security risk. The critical issue is the architecture behind it.
Business & Investments
Nvidia Brings in Wall Street: What Its $500 Billion Bet Really Funds
Nvidia wants to mobilize more than $500 billion for AI infrastructure. The key question is not only the amount, but who will carry data center risk.
Regulation & Privacy
Claude Marks AI Text: Why a Watermark Is Not a Truth Stamp
Anthropic is adding machine-readable markings to Claude output. What that means for the EU AI Act, detection, and editorial responsibility.
AI News
GPT-5.6-Cyber: OpenAI hands defenders a model that writes exploits
OpenAI released GPT-5.6-Cyber for security research. It found two unknown Chrome flaws, and access requires identity verification and a hardware key.