Five items that matter to a small business, with sources. News moves fast and vendor announcements are vendor claims — check the linked source before acting on any of it.
Jensen Huang has resigned as president and chief executive of Nvidia, effective immediately, the company said on Thursday, November 12 — ending 33 years at the helm of the company he co-founded. In a filing with the US Securities and Exchange Commission, Nvidia said its board accepted Huang's resignation on Wednesday, November 11. Chief financial officer Colette Kress, who has been Nvidia's finance chief since 2013, takes over as interim chief executive while keeping her CFO role, and the board named director Harvey C. Jones — a board member since 1993 — as its first-ever chair. The filing gives no reason for the resignation. It follows the board's announcement on Sunday, November 8, that it had opened an independent investigation into reports that Huang had been the target of a foreign intelligence operation, and reports of an affair with a woman alleged to have links to Chinese intelligence. Nvidia shares fell sharply on the news, and the company confirmed it will still report third-quarter results on Tuesday, November 17, as scheduled.
Why it matters: Nvidia is the company the whole AI buildout runs on — its chips, its software, its roadmap. A sudden change at the top, days before earnings and weeks into an investigation nobody outside the board has seen, puts the most-watched company in the AI economy into caretaker hands. For a small business, nothing changes today — the chips and the prices are the same — but the leadership question now hangs over every AI supplier's plans for next year.
TechCrunch: Nvidia's Jensen Huang resigns, chipmaker names CFO Colette Kress as interim CEO →
Anthropic's initial public offering has drawn about six times the orders for the shares available, Bloomberg News reported on Thursday, November 12. Bankers told prospective investors the company will prioritise its existing backers, leaving some would-be buyers with smaller allocations than expected — though the report stressed that details of the offering, including demand, could still change. The company and some existing shareholders aim to raise up to $64.4 billion by selling 825 million shares at $70 to $78 each, valuing Anthropic at up to about $1.4 trillion on a fully diluted basis, with a further $26 billion in a concurrent private placement. Shares are expected to start trading on the Nasdaq under the ticker "ANT" on November 19, with pricing expected on November 18. Regulatory filings put Anthropic's annualised revenue run rate at $81 billion as of October 31 — up from $71 billion at the end of September and $65 billion at the end of July.
Why it matters: Six times oversubscribed is a snapshot of demand, not a price — order books move until pricing day. But it sets up the largest listing in history as a public vote on the whole AI trade, a week after the Fed warned that AI returns may not justify AI valuations. For a small business, the listing matters less than the run rate behind it: $81 billion a year and rising fast says the AI labs' customers are spending at a scale no small-business budget can ignore.
Reuters: Anthropic's IPO is about six times oversubscribed, Bloomberg News reports →
Google introduced Gemini 4 Hydrogen on Thursday, November 12 — a faster, cheaper model built for everyday coding and agentic work, and the third member of the Gemini 4 family alongside Carbon and Argon. Google says Hydrogen finishes tasks about 2.5 times faster than Gemini 3.8 Flash, the model it replaces, and prices it flat at $1.50 per million input tokens and $7.50 per million output tokens across its one-million-token context window. It is rolling out across Google AI Studio, the Gemini API, Gemini for business, the Gemini app and Google Search, and to paid Google AI Pro and Ultra subscribers in the Antigravity coding environment. Free Gemini users get it automatically when the app decides it suits the prompt. The launch comes two days after Gemini 4 Carbon, and Google is now shipping new models at a pace that makes last month's flagship look dated.
Why it matters: Hydrogen is the model most small businesses will actually meet — it is the one inside Search and the free Gemini app, and at $1.50 per million input tokens it is the cheapest way into Google's newest generation. The practical point is the cadence: a model family that gains a member every few days means any workflow built around a named model needs a plan for when that model is replaced.
Google: Introducing Gemini 4 Hydrogen for Agentic Workflows →
New York opened registration under its RAISE Act on Tuesday, November 10, starting the clock on the first state AI safety law aimed squarely at the biggest labs. Large frontier developers — companies with more than $500 million in annual revenue, a category TechCrunch names as including OpenAI and Anthropic — must register with the state by January 1, 2027, when the law takes effect. Once registered they must publish safety frameworks, report critical safety incidents within 72 hours, and issue public transparency reports for their models. Fines run up to $1 million for a first violation and $3 million for later ones. The law is administered by a new Office of Digital Innovation, Governance, Integrity, and Trust (DIGIT) inside the Department of Financial Services. Governor Kathy Hochul also launched the New York Center for AI Responsibility and Research at Binghamton University, backed by $55 million, to independently evaluate advanced systems.
Why it matters: This is the first hard deadline the frontier labs have to meet that they did not set themselves — register, publish, report, or pay. It lands alongside California's SB 53 and ahead of Illinois's SB 315, so the compliance map for AI vendors is now being drawn state by state. For a small business, the direct effect is nil — the revenue threshold keeps you out — but the indirect one is real: the safety frameworks and incident reports the labs must now publish will become the paperwork your own AI suppliers hand you.
Securities.io: New York Opens RAISE Act Registration for Large Frontier AI Developers →
LMSYS Org launched the SGLang Consortium on Wednesday, November 12, at the inaugural SGLang Summit in San Francisco — a rare alliance of rivals to steer the open-source software that runs AI models once they are trained. Founding members are Nvidia, AMD, Intel, Google, Meta, Cloudflare, Modal and RadixArk, who will work with SGLang's maintainers on the project's roadmap, focused on frontier model support, reasoning and multimodal workloads, distributed inference and hardware optimisation. SGLang is the open-source inference engine, part of the PyTorch ecosystem, that already powers large production deployments at Meta, Microsoft, Google, Amazon and others. Deloitte estimates inference — running models, rather than training them — will account for roughly two-thirds of all AI compute in 2026, up from about half in 2025.
Why it matters: The money in AI is shifting from training models to running them, and the running is where the costs land on everyone else's bill. A shared, open inference engine backed by every major chipmaker and cloud means a small business's AI tools can run on whichever hardware is cheapest, instead of being locked to one vendor's stack. It is also a quiet admission: even rivals agree that no one company should own the plumbing.
This day has a permanent page: November 13, 2026. Every day is kept — see the News archive. The News page always shows the latest day.
← Latest day · News archive · Previous day: November 12, 2026
For what the AI companies and their leaders are saying on social media — claims and opinions, not verified news — see The Social Wire →