Nvidia Loses Its Founder, Labs Get Sued for Slowing AI, and SMART Flies Into Trouble

Three days of AI news arrived with the weight of three years. The founder of the company that built the infrastructure underlying the AI boom resigned under pressure, four of the most powerful AI labs face an antitrust suit accusing them of illegally pumping the brakes, and the federal agency overseeing American airspace quietly rolled an AI system to every corner of the country before a nationwide ground stop.

Jensen Huang Resigns From Nvidia

On Thursday, November 13, Nvidia announced that Jensen Huang had resigned as CEO, effective immediately. CFO Colette Kress will serve as interim chief executive while the board searches for a permanent successor. Harvey Jones was simultaneously named Nvidia's first-ever board chair.

The company did not publicly explain the departure, but reporting from Fortune and CNBC tied it to Huang's admission of an extramarital affair with a woman described by U.S. intelligence officials as a Chinese spy. An 8-K filed with the Securities and Exchange Commission confirmed the board accepted the resignation Wednesday and placed Kress in the top role, noting she would retain her CFO duties during the transition.

The exit removes the most prominent public face of the AI chip industry at a moment of acute sensitivity. Nvidia's $500 billion AI financing initiative, which Huang had championed, is now in question. The Senate Banking Committee chair has summoned both Huang and Commerce Secretary Howard Lutnick and threatened subpoenas over the initiative's structure. Nvidia stock extended declines the day after the announcement, with Wall Street analyst Dan Ives estimating Kress could hold the interim role for roughly a year.

Kress has served as Nvidia's CFO since 2013. She was the company's longest-serving C-suite executive before Thursday, and while she has deep knowledge of its finances, she has never run product, engineering, or customer relationships at the CEO level. Huang, who co-founded Nvidia in 1993 and has been its only CEO, built it from a graphics chip company into the dominant supplier of AI training hardware. The company's market capitalization peaked above $3.5 trillion under his leadership.

Four AI Giants Sued Over an Alleged Slowdown Pact

A class-action lawsuit filed September 18 in the Northern District of California is gaining fresh attention this week. Four paid subscribers — led by plaintiff Nathan Buist — sued Anthropic, OpenAI, Google, and SpaceXAI, alleging the companies violated federal antitrust law by secretly agreeing to slow the pace of AI capability improvements.

The complaint traces the alleged agreement to a July 2026 public statement called "Pacing the Frontier," signed by representatives of all four companies along with a coalition of AI safety nonprofits. Plaintiffs argue the statement was not a good-faith policy position but a cover for an illegal agreement to reduce competition in the market for AI subscriptions. If the labs are racing less aggressively, paying subscribers get less value for their money, the suit contends.

The hook for the antitrust claim is a September 12 essay by Anthropic CEO Dario Amodei titled "We Must Pace the Frontier," which called explicitly for coordinated limits on the rate of AI progress. The plaintiffs cite it as evidence the coordination moved from public statements into active business policy.

The defendants deny the allegations. Each company has characterized its participation in the "Pacing the Frontier" statement as a good-faith contribution to AI safety policy, not a commercial agreement. The case is in its early stages; no trial date has been set. Bloomberg Law, Tom's Hardware, and the Daily Journal have reported on the suit's details.

LangChain Closes $300M Series C at $3B Valuation

LangChain announced Thursday that it has raised $300 million in a Series C round, more than doubling its valuation from $1.25 billion to $3 billion. The round was led by CapitalG, Google's growth equity fund, with participation from GIC, DFJ Growth, IVP, Sapphire Ventures, Salesforce Ventures, Insight Venture Partners, Sequoia, Amplify, and Benchmark.

The company says more than 7,000 organizations now use LangSmith, its agent observability and governance platform, to build and operate AI agents in production. The funding will go toward expanding LangSmith's capabilities, with new features expected on November 20.

LangChain's rise tracks the broader shift in enterprise AI from standalone language models to multi-step agent workflows. As companies deploy agents that take actions — not just answer questions — the need for tools to monitor, debug, and govern those agents has grown sharply. LangChain is betting its platform becomes the standard layer for that work, the same way GitHub became the standard layer for code.

The round brings total funding to $460 million. LangChain did not disclose revenue or profitability figures.

FAA's AI Air Traffic Tool Flew Into Every Center Before the Ground Stop

A Washington Times report published November 13 revealed that the FAA expanded its SMART AI-assisted air traffic management system to all 22 of its Air Route Traffic Control Centers in the days leading up to a nationwide ground stop. The timing has renewed scrutiny of a system that House lawmakers had already asked the agency to pause.

SMART — which stands for System for Multi-center Air traffic management Research and Technology — launched in the Washington, D.C. area in September, combining roughly 200 data streams including weather forecasts, flight paths, air traffic load, and controller staffing levels into advisory recommendations for traffic managers. The FAA stressed from the start that the system is advisory: its outputs require human review before action, and it does not control aircraft.

In October, more than two dozen House lawmakers wrote to the FAA asking it to pause SMART at the initial three airports pending a fuller safety review. That request appears not to have halted the nationwide expansion.

The company behind SMART has said its system did not cause the ground stop, and the FAA has not publicly established what did. But the juxtaposition — a nationwide AI deployment three days before a nationwide disruption — has handed critics of AI in critical infrastructure a concrete data point to press.

The episode lands as regulators across transportation, healthcare, and finance are grappling with a common problem: AI tools in advisory roles can be difficult to hold accountable when things go wrong, because no single decision point is clearly automated. The FAA situation is likely to deepen that debate in aviation specifically, where the margin for error is narrow and public trust is essential.