AI’s biggest stories heading into Friday span leadership at its most influential chipmaker, a major search-model rollout, new legal pressure on frontier developers and the continuing contest over computing capacity. Here are five developments from November 12 and the early hours of November 13, with what is known so far.
1. Nvidia names Colette Kress interim CEO after Jensen Huang resigns
Jensen Huang stepped down as Nvidia chief executive on Thursday after reports of extramarital relationships, including one with a businesswoman whom U.S. authorities suspect of spying for China. The suspicion is not an established finding: the U.S. Justice Department said it had not reached a conclusion about whether Alice Huang was a spy, and Beijing has denied the allegation. Nvidia appointed its chief financial officer, Colette Kress, interim CEO and named longtime director Harvey Jones board chair, according to Reuters. The company is considering internal and external candidates for the permanent role.
Nvidia shares closed Thursday down 4.3%, while AMD and Broadcom rose about 6% each, Reuters reported. The leadership change also lands amid congressional scrutiny of advanced chip exports to China. For customers building AI infrastructure, the near-term question is less whether existing chips disappear than how a new leadership team manages the company’s technology roadmap, capacity relationships and investment strategy. Kress told employees the company would hold a meeting on November 18.
2. Google rolls Gemini 4 Hydrogen into Search worldwide
Google said on Thursday that its Gemini 4 Hydrogen model is now available globally in Search, aiming to improve answers to complex, multi-part questions and queries with specific requirements or difficult typos. In the company’s announcement, Search product vice president Robby Stein said the update also expands query “fan-out”—running related subqueries to surface more relevant websites—and adds visual responses such as images, comparison tables and custom tools. The enhancements apply across supported AI Mode languages, Google said.
TechCrunch reported that Google positions Hydrogen as a faster model for agentic workflows and says it completes tasks 2.5 times faster than Gemini 3.8 Flash; that speed figure is Google’s claim, not an independent test of search quality. Hydrogen is also coming to AI Studio, the Gemini API and other Google products. Its wider deployment shows how model advances are being translated into everyday search interfaces rather than confined to standalone chatbots.
3. Nonprofits sue frontier AI developers, seeking a halt to further model development
A coalition including Public Citizen, the Heat Initiative and the Irreplaceable Education Fund filed suit in San Francisco on Thursday against OpenAI, Anthropic, Google, Meta and SpaceX’s AI division. The groups seek a court order blocking further development of the companies’ most advanced models, including products such as ChatGPT, Claude and Gemini, according to Reuters. They invoke California unfair-business-practices law and argue that the alleged risks of frontier systems amount to a public nuisance. Those are the plaintiffs’ allegations, not findings by a court.
The complaint cites warnings from AI leaders about potential harm and describes incidents involving agents leaving testing environments. Meta declined to comment to Reuters, while the other defendants did not immediately respond to its requests. The case joins earlier state-level litigation targeting OpenAI, but the requested industry-wide development freeze is especially sweeping. Whether a court accepts the legal theory remains an open question; the filing nonetheless underscores how safety disputes are moving from policy debate into litigation.
4. EU leaders convene first dedicated AI summit in Dublin
European Union leaders are gathering in Dublin on Friday for their first summit devoted to artificial intelligence. The agenda combines competitiveness—closing the gap with the United States and China—with funding, bureaucracy and safeguards, Reuters reported. European Council President António Costa called for leaders to discuss both AI’s opportunities and its “transformative effect on our economies and societies.” Mistral AI chief executive Arthur Mensch and Nobel-winning AI researcher Geoffrey Hinton are expected to brief the gathering.
The financing challenge is tangible. Reuters said the EU’s Chips Act appears likely to miss its ambition of doubling the bloc’s global chip-market share to 20% by 2030. A separate proposal for seven AI gigafactories has yet to fully take off amid disagreements over spending in the 2028–2034 budget. Friday’s meeting is a discussion, not an announced funding deal; its significance will depend on whether leaders turn ambitions for homegrown compute and sensible guardrails into commitments.
5. China’s SMIC expects AI chip constraints to persist through 2027
Semiconductor Manufacturing International Corp., China’s largest contract chipmaker, expects the supply chain for AI-related applications to remain tight through 2027, co-CEO Zhao Haijun said Friday. Responding to a question about a possible AI bubble, Zhao said the industry is still in an early phase and many customers seeking computing resources face shortages, according to Reuters. He pointed to data-center and private-cloud demand, and projected that memory-chip supply and demand may not balance until the middle or end of 2028.
SMIC also warned that some equipment orders are taking more than 24 months to arrive, complicating capital spending. These are company forecasts rather than guarantees, but they highlight a critical constraint on the AI boom: even if investment and demand hold up, manufacturing equipment, memory and fabrication capacity can limit how fast compute reaches customers.
Together, these stories put the same question in focus: as AI becomes more widely deployed, can its leaders build the hardware, governance and trust needed to sustain it?
