Good morning from the wAIve team in London.

Today, Friday 13 November 2026: It’s all about Jensen Huang today. Either you believe he had to go or it’s the biggest own goal in corporate history. Is the business bigger than the man? We’re about to find out. NHS England has quietly deleted its claim that 137 hospital trusts were benefiting from Palantir. Meanwhile, Nebius plans to spend $40 billion on AI infrastructure, largely with its customers' money.

  BREAKING  

Jensen Huang has resigned as CEO and president of Nvidia, effective immediately, following an espionage scandal involving an alleged affair with a Singaporean executive believed by US intelligence to be a Chinese spy. CFO Colette Kress has been named interim chief executive. Democrats have launched a formal congressional investigation, Huang has agreed to testify before Congress, and the scandal has erased roughly $1.4 trillion in market cap. Reports of a third affair involving a current Nvidia employee emerged on the same day as the resignation.

THE PLAY: Huang wasn't an ordinary chief executive. He was the architect of the firm's chip strategy, its key relationship with US government officials, and its advocacy for open-source AI. No succession plan exists. Nvidia isn't going anywhere, it's too big to sink, but, but, but... sensible minds will audit dependency on Nvidia hardware.

DEEPER READ: Who is Colette Kress, Nvidia's new interim CEO?

  POLICY  

Australia's government has introduced legislation that would, for the first time, enshrine mandatory safety requirements for frontier AI companies in domestic law. Under the proposed bill, a new regulator would have authority to compel independent model audits of the most capable AI systems and impose financial penalties for non-compliance. Canberra has already stated it intends to establish domestic rules before entering international negotiations.

This follows Australia's earlier mandatory AI incident reporting requirements. The pattern is clear: Canberra is building a layered regulatory architecture that will impose real compliance obligations on any AI company operating in or selling to Australian markets.

THE PLAY: For global businesses, this means a third set of laws alongside the EU AI Act and emerging US federal proposals. Each imposes different audit, disclosure, and testing obligations.

  EXCLUSIVE  

NHS England has removed a claim from its website that 137 hospital trusts were actively benefiting from Palantir's AI-powered data platform, after the Guardian revealed that 29 of those trusts said they were either not using the system at all or not using most of its applications. The original claim described all 137 trusts as being both "live with" and "realising benefits from" the technology.

THE PLAY: Palantir delivers results but for some, especially within the Labour govt., that comes at a price. But when procurement decisions are built on inaccurate figures, questions need to be asked.

Cloud AI provider Nebius has announced plans to spend $40 billion expanding its AI infrastructure, funding the commitment primarily through customer prepayments and loans rather than equity. The announcement comes as delays at a Microsoft data centre have raised questions about the financial model underpinning its expansion.

Why it matters: Nebius is effectively transferring capital risk to customers who prepay for capacity that does not yet exist. As wAIve has previously noted, CoreWeave has adopted similar structures. Any business entering long-term AI compute contracts should scrutinise what happens to their prepayments if delivery timelines slip.

Investment firm Norwest has closed a $4 billion fund, the largest in its history, with AI as the stated primary focus. The fund will deploy into companies across AI infrastructure, enterprise software, and AI-enabled services.

Why it matters: For businesses evaluating AI partnerships or acquisitions, this level of capital arguably indicates that the AI show is far from over; the sector will continue to rise through 2027.

UK GDP grew by 0.7% in the third quarter of 2026, the fastest expansion in more than two years. Economists and the government cited AI-driven productivity gains as a contributing factor. The growth puts the UK on course to be the second-fastest growing G7 economy in 2026.

The regional angle: For UK-based business leaders, this is the first meaningful macroeconomic data point linking AI adoption to measurable national output growth. It strengthens the case for board-level investment in AI productivity tools, and training.

Closing prices, Thursday 12 November (US markets). Not investment advice.

An AI participant that joins your Google Meet calls in real time, generates meeting notes, creates whiteboards, and produces structured documents during the session. Available now in beta on Team and Enterprise plans; notes and transcripts are free during the beta, and documents and whiteboards made in a call are free until 10 February 2027, with no price announced after that.

THE PLAY: Claude is coming for the virtual meeting space. Add Claude Meetings to your next strategy session and ask it to produce a structured action-item list with owners and deadlines before the call ends.

A real-time speech recognition tool that handles background noise and cross-talk without losing accuracy, designed for voice agents and live call environments. Benchmarks show it narrowly leads competitors on accuracy, though it costs roughly twice as much per minute as Microsoft's equivalent and runs more slowly.

THE PLAY: Use Scribe v3 Live to transcribe your next client call or external meeting.

Nvidia's implosion this week is dominating coverage for the right reasons. But the deeper lesson for business leaders is not about one man's personal failings. It is about what happens when critical infrastructure depends on a single supplier, a single personality, and a single geopolitical relationship, all simultaneously.

The same week the chip giant lost its chief executive, two separate stories emerged that illustrate the same underlying problem from different angles. Neither received the attention it deserved.

A cyberattack on Fireworks AI, an AI infrastructure provider, allowed hackers to access chat transcripts belonging to Notion customers. This is Fireworks' second significant security incident in as many months. Neither company is a household name (although Notion is a tech darling for some). But they both sit inside the AI supply chains of thousands of businesses whose customers and employees had no idea their data was passing through a third-party AI layer.

The Notion breach was caused by a supplier's supplier's failure. Most organisations cannot name their AI supply chain beyond the first layer. This is a major governance gap, and regulators are beginning to notice.

Separately, OpenAI's safety team has published research showing that frontier AI models, including its own most advanced systems, are increasingly concealing their internal reasoning. As developers tune models for speed and cost efficiency, the chain of reasoning that makes AI decisions auditable becomes obscured.

Three actions matter immediately. First, map your AI supply chain beyond the first layer. Know which infrastructure providers your AI tools depend on and what their security posture is.

Second, ask your AI vendors whether their models produce auditable reasoning trails for any decision that affects a customer or employee.

Third, treat AI suppliers the same as any other firm but with added red flag systems. This is about operational continuity and regulatory readiness.

The crisis at Nvidia is dramatic because one individual was at the centre of everything. But the structural lesson applies at every level of the AI stack.

Most people prompt AI with the task but forget to specify the audience. The result is output that is accurate but pitched at the wrong level: too technical for a board, too vague for a specialist, too long for someone reading on a phone. Anchoring your prompt to a specific, named reader type transforms the output immediately.

It works because it forces the model to filter every sentence through a specific reader's priorities rather than producing a generic summary. The CFO version and the CTO version of the same briefing should look entirely different, and with this prompt, they will.

Thanks for reading, from the wAIve Daily Team. Found this useful? Forward it to someone who needs it. Tell us what you need to know. Reply with your questions. And send us a story if you have one.

wAIve Daily is published with Centai Strategy