AI news, 14 November: OpenAI’s ChatGPT Wallet lets agents shop for you
OpenAI launches ChatGPT Wallet for agent purchases. OpenAI rolled out ChatGPT Wallet on 13 November, enabling its “dots” AI agents to complete online purchases on behalf of users after explicit approval. Built with American Express, Mastercard and Visa, the system uses payment tokens so neither the agent nor the merchant sees the user’s actual card details; each transaction requires a biometric check such as Face ID. It is initially available to U.S. ChatGPT Pro subscribers aged 18 and over with eligible cards. According to OpenAI and reports from Reuters and TechCrunch, the company is not taking a cut of transactions at launch and is providing guidance so businesses can accept agent-led checkouts without breaking existing systems. The move marks a concrete step toward everyday agentic commerce while keeping users in control of every payment.
Gemini connects to Google Wallet for spending insights. Google began rolling out an optional link between the Gemini app and Google Wallet for U.S. users, letting the assistant answer personal-finance questions based on real spending data. Users can tag @Google Wallet to review past outlays, spot forgotten subscriptions, generate budget charts or plan holiday spending. Deeper insights are available by securely linking bank and card accounts via Plaid. Google stresses the connection is read-only: Gemini cannot move money, make payments or manage accounts, and access can be revoked at any time. The feature expands Gemini’s Personal Intelligence tools and arrives alongside OpenAI’s payment capability, highlighting how major AI assistants are embedding deeper into daily financial life.
Nonprofits sue to freeze frontier AI development. A coalition including Public Citizen, the Alliance for Secure AI and others filed suit in San Francisco Superior Court against OpenAI, Anthropic, Google, Meta and SpaceX’s AI unit. They seek an injunction barring development or deployment of models more capable than the companies’ current systems, citing California’s unfair-competition law and public-nuisance rules. The complaint points to the labs’ own warnings about catastrophic risks, alleged incidents of AI systems escaping test environments, and potential for cyberattacks or bioweapons. The groups argue companies should not self-regulate products they admit could cause extreme harm. The case, reported widely on 12–13 November, adds to public debate over whether courts should intervene before more powerful systems arrive.
Google releases Gemini 4 Hydrogen. Google launched Gemini 4 Hydrogen, a faster and lower-cost model in its Gemini 4 family aimed at coding and agent workloads. Priced at roughly $1.50 per million input tokens and $7.50 per million output tokens with a 1-million-token context window, it scored strongly on independent intelligence indexes and is available via the Gemini API, AI Studio, Search and other Google products. The release replaces earlier Flash variants and arrives as Google also adjusts pricing on higher-tier models later this month.
SpaceX commits more than $74 billion to AI campuses. SpaceX announced plans to invest over $74 billion in two AI supercomputing campuses—one at its Starbase site in Texas (more than $24 billion, with on-site power generation) and a larger facility in Cochise County, Arizona (about $50 billion). The Arizona project still needs local approvals and is projected to create thousands of jobs; the Texas site would generate its own electricity. Reuters and Bloomberg reported the plans, underscoring the massive infrastructure race fueling advanced AI training and inference.
Nvidia succession update. In a fresh development since Jensen Huang’s resignation, the Financial Times reported that Nvidia’s board has approached Arm CEO Rene Haas, a former Nvidia executive, as one possible permanent successor. Nvidia stated it is still selecting a search firm and has not formally engaged any candidates, calling speculation premature. CFO Colette Kress remains interim CEO.