Good morning from the wAIve team in London.

Today, Thursday 12 November 2026: A week for finding out who is really in charge. Nvidia's board met to decide whether to oust Jensen Huang as Congress called him in to explain himself, and New York started making the biggest AI labs register under its new safety law. Meanwhile, Qantas is cutting 850 jobs and giving AI the credit.

  BREAKING  

Nvidia's board was reportedly ready to vote on removing founder and CEO Jensen Huang as early as Wednesday evening, according to multiple reports. Congressional Republicans have summoned Huang for a closed-door briefing, while the FBI investigation has broadened. Microsoft has overtaken Nvidia in market cap for the first time since mid-2025.

THE PLAY: When they say AI is coming for your job, this isn't what they had in mind!

Analysts describe Huang as irreplaceable in the short term: he is Nvidia's strategy, its relationships, and its credibility with customers. As Arianna Huffington put it in a tweet:

"The Jensen Huang saga continues to unfold (and don't you feel like we're in act one of a five-act play?). But one question has already been definitively answered: what the biggest limiting factor of the AI revolution will be. As this story makes clear, it's not going to be compute or energy or Nvidia's GPUs. It's going to be what it's always been: human nature."

  POLICY  

New York's RAISE Act took effect this week, requiring major AI developers including OpenAI and Anthropic to register with state authorities, disclose their safety practices, and report critical incidents or face financial penalties. The law is the first state-level AI safety statute in the US with genuine enforcement power, and it sets a precedent other states are watching closely.

THE PLAY: If your vendors are now subject to mandatory safety disclosure and incident reporting, it's reasonable to request that info as part of supplier due diligence.

  INDUSTRY  

Qantas has announced it will cut 850 Australian jobs and outsource work to India through a partnership with Accenture, citing "unprecedented" technological disruption driven by artificial intelligence. It is the third major AI-linked corporate restructuring in Australia in three weeks, following similar moves from Canva and National Australia Bank, which announced cuts of 2,500 roles.

THE PLAY: Across industries, AI-linked restructuring is accelerating. And Qantas is talking about AI capability, knowledge work, not just cost and ROI.

UK AI chip startup Fractile has secured a $720 million funding round, making it one of the best-funded alternatives to Nvidia in the race for purpose-built AI chips. The company designs chips optimised for running AI models efficiently rather than training them, targeting the growing inference market where businesses deploy AI at scale.

Why it matters: With storm clouds over the Nvidia boardroom, this is a credible domestic alternative. Watch this space for procurement implications within 18 months.

CoreWeave, a major AI cloud provider, is increasingly requiring customers to pay upfront for capacity, borrowing against specific contracts and offering shorter-term deals. Its third-quarter earnings revealed the shift as it seeks to fund rapid expansion without absorbing all the financial risk itself.

Why it matters: Shorter contracts may mean higher prices for the same capacity. If you're negotiating AI cloud capacity now or in the next six months, market leverage has moved back toward the provider.

European defence AI company Helsing has launched Clarity, a platform that distils battlefield sensor data into weapons targeting recommendations, with human authorisation required at every decision point. The company positions it as a European-controlled alternative to US systems, with every decision carrying a named, accountable human.

The regional angle: European defence spending on AI is accelerating sharply, and Helsing is positioning itself at the centre of that market.

Closing prices, Wednesday 11 November (US markets). Not investment advice.

An AI agent that answers plain-language questions about your organisation's data and surfaces honest operational intelligence directly to leadership, without requiring analysts to build custom reports.

THE PLAY: Ask your data team to connect Genie to your sales or operational data, then spend 20 minutes asking it the questions you would normally have to wait two weeks for a report to answer.

A free AI research assistant that lets you upload documents, reports, and PDFs and then ask questions, generate summaries, or identify contradictions across the material. Best for executives who need to absorb large volumes of text quickly. Free at notebooklm.google.com.

THE PLAY: Upload your docs and ask NotebookLM to identify the top five assumptions your strategy relies on not yet validated by evidence. If you haven't played around with NotebookLM yet, it's a game changer.

A new report from England's police inspectorate has lessons for anyone responsible for people: hiring, credit, fraud detection, insurance pricing, or customer screening.

The watchdog found that officers were treating facial recognition matches as evidence rather than as leads requiring further investigation.

Innocent people, disproportionately from ethnic minority groups, faced wrongful arrest. Some officers acted on potential matches without making sufficient independent checks. The technology worked as designed. The system failed at the point where machines dictated human decisions.

This is the pattern that should concern every C-suite. AI systems marketed as highly accurate are often accurate on average, but inconsistent at the individual level.

The facial recognition report is not unusual. The AI produces a result and someone acts on it. The gap is where human intelligence (not artificial) should intervene.

New York's RAISE Act, which came into force this week, signals where regulation is heading. Mandatory incident reporting, safety disclosures, and audit requirements are becoming standard expectations.

The EU AI Act already classifies biometric identification, credit scoring, and employment-related AI as high-risk systems requiring human oversight and audit trails. Businesses that can't show human oversight for key decision making are leaving themselves exposed.

Before deploying any AI system that influences decisions about individuals, map every point where machine output becomes a consequential action. Assign a named person accountable for that decision.

Before walking into a meeting on a subject you do not know well, use AI to compress hours of reading into ten minutes of targeted preparation. This is not about generating a summary. It is about identifying exactly what you need to know to participate meaningfully and ask the right questions.

This works because it forces the AI to prioritise rather than summarise, delivering the minimum viable knowledge rather than a comprehensive overview you will not retain before you reach the door.

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