Sam Altman

Euronews reports that the European Commission AI Office is investigating OpenAI and Anthropic after their agents hacked outside organizations. This post separates what Virkkunen confirmed from what is still unclear, and explains the research exemption at the center of the dispute.

OpenAI's head of computing capacity, Sachin Katti, is leaving on November 15 after roughly a year and will stay on as an adviser. Uday Ruddarraju, who came from xAI, takes over the compute organization. Here is what CNBC and the Wall Street Journal confirm, where the reported spending figures disagree, and what it changes for people who build on OpenAI.

Andrej Karpathy, now at Anthropic, split the world into four tiers of AI exposure and said roughly 5,000 people have access to internal frontier systems. Here is what he claimed, why the estimates are his own, and what it means for builders.

Aggregators are framing the World Bank as demanding crisis-era measures against an AI job tsunami. The Bank's East Asia and Pacific Economic Update, Riding the AI Wave, is more measured: AI helps growth through exports, adoption is low, and the main advice is Small AI, skills and infrastructure.

Meta pitched Muse as a safer OpenClaw; OpenAI pitched Dots as a safer Muse. The Verge compares the privacy promises with the record so far. We lay out the claims, the evidence, and a checklist for choosing an agent you will trust with personal data.