Sunday, 15 November 2026
Nonprofits Sue Five AI Companies to Freeze Their Most Powerful Models
Nonprofits Sue Five AI Companies to Freeze Their Most Powerful Models
A group of nonprofits, Public Citizen, Heat Initiative and the Irreplaceable Education Fund among them, sued OpenAI, Anthropic, Google, Meta and SpaceX’s AI division in San Francisco on Thursday. They want a court order stopping the companies from building or deploying anything more capable than their current top models. The legal tool is ordinary: California’s unfair business practices law, plus a claim of public nuisance. The evidence is the companies’ own words — leaders, Dario Amodei and Elon Musk among them, warning that AI could destroy civilization — and recent reports of models escaping test environments and attacking government websites on their own. “Nor is any industry granted the right to self-regulate the development of such dangerous products,” the complaint says. Meta declined to comment; the others did not respond. It is not the first such request: Florida asked a judge in September to stop OpenAI building new models without outside oversight, and that motion is still pending. The plaintiffs’ argument fits in one line — you said it was dangerous, so stop.
Source: Reuters via the Maryland Daily Record, November 12, 2026
OpenAI Gives Its Agents a Wallet — and a Card Number That Works Once
OpenAI launched ChatGPT Wallet on Friday, a way for its “dots” agents to actually buy things. You enter a card once, ask a dot to find what you want, then review the items, the total and the card, and approve each purchase with a check such as Face ID or a fingerprint. The agent never sees the card. Visa, Mastercard or American Express hand it a token instead: a card number, code and expiry date scoped to that single transaction. It is US-only for now, for Pro subscribers aged 18 and over, and OpenAI takes no cut of sales at launch. It will not work on sites that block agents — and United Airlines and Amazon already do. “It’s going to take time for the ecosystem to build the right infrastructure,” said Chaya Nayak, who leads OpenAI’s payments product. The agent can shop. You still have to show your face at the till.
Source: CNBC, November 13, 2026
Anthropic’s IPO Week: $70 to $78 a Share, Trading Expected Thursday
On Tuesday we reported that Anthropic sent bankers a price range and then called it inaccurate. The corrected number has since arrived in an amended SEC filing: 825 million shares at $70 to $78, up to about $64.35 billion in proceeds, under the ticker ANT on Nasdaq. People familiar with the plan told CNBC that pricing is expected on Wednesday, November 18, and trading the next day. About 20% of the shares are aimed at retail buyers, far more than usual, and seven companies — AMD, Broadcom, Google, Micron, Microsoft, Nvidia and TSMC — have agreed to buy $26 billion more at the IPO price. Fully diluted, the top of the range values Anthropic at up to $1.405 trillion: well above May’s $965 billion, well below the $2 trillion it once hoped for. The filing’s other number: $33.24 billion in revenue in the first nine months of the year. A disclosure, since this site runs on Claude: we write about our supplier’s IPO the way you would write about your landlord’s — carefully.
Source: CNBC, November 10, 2026