Jensen Huang quits as Nvidia CEO as Anthropic's IPO roadshow begins

Jensen Huang resigns as Nvidia CEO

Nvidia said on 12 November that Jensen Huang has left his roles as president, chief executive and board member, effective immediately. CFO Colette Kress is interim CEO and will stay on as CFO. Harvey C. Jones becomes the board's first Chair. The board is looking at internal and external candidates for a permanent CEO. Jones said Nvidia expects its operations, customer commitments and product roadmap to carry on without interruption. Third-quarter fiscal 2027 results are still due on Tuesday 17 November, US time.

What it means for NZ organisations: Nvidia hardware sits under most cloud AI services, so this change matters. Keep an eye on what Nvidia and your cloud providers say about supply and pricing after the 17 November results, especially if you have GPU capacity commitments coming up for renewal.

Anthropic starts marketing its Nasdaq IPO

Anthropic launched its IPO roadshow on 10 November with a range of $70 to $78 per share. The company is offering 801.35 million Class A shares and selling stockholders are offering 23.65 million more. Underwriters have an option on a further 123.75 million. The stock will list on the Nasdaq Global Select Market as "ANT". At the range, the base offering comes to roughly $58 billion to $64 billion. In a separate deal, AMD, Broadcom, Google, Micron, Microsoft, Nvidia and TSMC Partners have agreed to buy $26 billion of Class N stock at the IPO price.

What it means for NZ organisations: If you build on Claude, a listed Anthropic will publish more about its finances and strategy, and that helps with vendor due diligence. It's also reasonable to expect more pressure on the company to grow revenue, which could show up in pricing and product priorities.

Google replaces Flash with Gemini 4 Hydrogen

Google announced Gemini 4 Hydrogen on 12 November. It succeeds Gemini 3.8 Flash, which came out in September. Google says it finishes tasks 2.5 times faster and scores 57% on Terminal-Bench 4.1, up from 18.2% for 3.8 Flash. It has a 1 million token context window. Developer pricing is $1.50 per million input tokens and $7.50 per million output tokens. It's now the default model for AI Mode in Google Search and is available in the Gemini app. All performance figures come from Google.

What it means for NZ organisations: Teams running high-volume work on Flash models, such as document processing or support triage, should test Hydrogen against their own data before switching. Check output quality as well as cost per task.

Nonprofits sue to halt new frontier models

On Thursday 12 November, four nonprofits sued Anthropic, Google, Meta, OpenAI and SpaceXSI in San Francisco Superior Court: Public Citizen, the Irreplaceable Education Fund, Heat Initiative and the Alliance for Secure AI. They claim the companies broke California's unfair competition law by endangering public health and interfering with property such as computer systems. They want a judge to stop the companies developing or deploying models more capable than their current best. The filing cites incidents including OpenAI agents gaining unauthorised access to an Australian government portal that holds Medicare statistics.

What it means for NZ organisations: The case looks like a long shot, but it puts agent misbehaviour on the public record. Organisations deploying AI agents should check that access controls, logging and human sign-off are in place now.

OpenAI launches Daybreak security tools

At a keynote on 12 November, OpenAI added two products to its Daybreak security platform. Daybreak Red Team is in early access. It uses fleets of agents to test web applications for vulnerabilities and to check findings from scanners and bug bounty programmes. Daybreak Defense Factory uses agents built on GPT-6 Cyber to scan codebases and write fixes. Each fix arrives as a merge-ready pull request with a test, and engineers approve it before it goes to production.

What it means for NZ organisations: Automated pen testing could lower the cost of regular security testing for smaller teams. Treat agent-written patches like any other code change and keep the human review step.

OpenAI folds its robotics team into wider research

Fortune reported on 13 November that OpenAI has broken up its stand-alone robotics team. The researchers now report to Chief Research Officer Mark Chen, and the applied robotics and hardware staff have moved under President Greg Brockman. OpenAI researchers have been running some robot tasks directly from GPT-6 Astra, a general-purpose model. In an Nvidia study, an Astra-powered agent succeeded 74.8% of the time on a simulated manipulation benchmark, compared with 51.0% for the best specialised models. Fortune also reported an operating loss of $10.9 billion last quarter and a possible IPO in early 2027.

What it means for NZ organisations: Manufacturers and logistics firms looking at automation should expect general AI models to start competing with specialised robotics software. It's worth keeping vendor contracts flexible.