AI News, Week 46: Nvidia Earnings Without Jensen Huang
Also: AI agents pay by card, and Germans want an "emergency stop" for AI
The Week in 30 Seconds
Earnings Without Huang: Nvidia will report its quarterly earnings on November 17. Interim CEO Colette Kress will present the results, as Jensen Huang has resigned following reports of extramarital affairs.
Agents at the Checkout: OpenAI allows agents to pay with Visa, Mastercard, and American Express in ChatGPT, while United Airlines is blocking third-party agents. The EU wants to ensure consumers have control over such purchases.
Emergency Shut-Off Called for: According to Bitkom, 92 percent of Germans want an emergency shut-off switch for all AI systems. That same week, Geoffrey Hinton spoke to EU heads of state and government about the risks of AI.
This week’s AI news (Week 46) begins with an unusual scene: Nvidia releases its quarterly earnings, presented by interim CEO Colette Kress instead of Jensen Huang. In addition, users can now pay AI agents in ChatGPT, and a Bitkom survey reveals the extent of concerns about uncontrollable AI.
These news stories address issues that many companies are currently grappling with: How dependent are you on individual providers, and who is authorized to act on your behalf? Here, you can read about what’s behind the three most important AI news stories of calendar week 46 and what they mean for your company.
Nvidia Releases Quarterly Results; Colette Kress Presents Them Following Huang's Resignation
“
NVIDIA’s chips are still rolling, and so are the supply contracts. Politics is the bigger risk: If Washington tightens export regulations, it will affect prices and delivery times worldwide. Anyone planning AI infrastructure should not rely on a single chip manufacturer.
Robin Reuschel
Expert in AI Strategy & Organization
“
NVIDIA’s chips are still rolling, and so are the supply contracts. Politics is the bigger risk: If Washington tightens export regulations, it will affect prices and delivery times worldwide. Anyone planning AI infrastructure should not rely on a single chip manufacturer.
Robin Reuschel
Expert in AI Strategy & Organization
What This Means for Businesses
- Identify dependencies: Determine which of your AI applications run on Nvidia hardware at cloud providers. This will help you see where price or delivery changes will affect you.
- Review commitments: Clarify with your cloud and data center partners what capacity and pricing commitments apply during bottlenecks.
- Test alternatives: Try out your models on infrastructure from other chip manufacturers, such as AMD or Google’s TPUs, before you need them.
Read about how closely the federal government and Nvidia plan to collaborate in the AI News for Week 42.
OpenAI Lets AI Agents Pay with Credit Cards; the EU Wants to Ensure Buyers Have Control
“
Agents become customers, whether you plan for it or not. Define which purchases an agent is allowed to initiate on its own and where human approval is required. You need this rule for your own agents and for your customers’ agents.
Robin Reuschel
Expert in AI Strategy & Organization
“
Agents become customers, whether you plan for it or not. Define which purchases an agent is allowed to initiate on its own and where human approval is required. You need this rule for your own agents and for your customers’ agents.
Robin Reuschel
Expert in AI Strategy & Organization
What This Means for Businesses
- Test your store: Check whether your online store allows purchases by agents or blocks them as fraud. Make an informed decision about which option you want.
- Set approval rules: Define spending limits for your own purchasing or booking agents and specify the cases in which a human must provide confirmation.
- Ensure human contact: The Digital Fairness Act provides for the right to a human contact person. Check whether your customer service chatbot offers a clear path to human employees.
Our article on Agentic AI demonstrates how agents can work under supervision within a company.
Bitkom: 92 Percent of Germans Want an Emergency Shutdown for AI; Geoffrey Hinton Warns of Loss of Control
“
The concerns raised in the survey are also shared by your staff. No one can build an emergency shutdown for all AI systems. But you can do this for your own agents: limit their permissions and configure each agent so that it can be shut down at any time.
Robin Reuschel
Expert in AI Strategy & Organization
“
The concerns raised in the survey are also shared by your staff. No one can build an emergency shutdown for all AI systems. But you can do this for your own agents: limit their permissions and configure each agent so that it can be shut down at any time.
Robin Reuschel
Expert in AI Strategy & Organization
What This Means for Businesses
- Address concerns: Openly discuss risks during training sessions. People who understand how a system works and where its limitations lie will use it more safely.
- Enable shutdown: Ensure that every AI agent can be stopped at any time. Test this process before the agent goes live.
- Putting the debate into context: Experts are divided on the issue. Provide executives with facts so they can answer questions within their teams objectively.
Our page on AI transformation and change management explains how to get employees on board with AI.
Do you want to use AI agents with clear rules?
We’ll work with you to determine what decisions your agents are allowed to make on their own and how your teams will manage them.
Sources
- Nvidia: NVIDIA Announces CEO Transition, November 12, 2026
- Yahoo Finance: Nvidia Live Updates: Following Jensen Huang’s Resignation as CEO, Allegations of a Third Affair Emerge, November 13, 2026
- Fox Business: Jensen Huang’s allies push Nvidia’s board to reinstate former CEO: report, November 15, 2026
- European Commission: Commission Strengthens Consumer Protection Laws, Making Them Fit for the Digital Age, November 11, 2026
- The Next Web: OpenAI’s ChatGPT Wallet lets its AI agents shop without seeing your card, November 13, 2026
- TechCrunch: United Airlines blocks AI agents from booking flights, citing cybersecurity risks, November 12, 2026
- Bitkom: Germans Want More AI Security, November 10, 2026
- Euronews: Ukraine envoy row hijacks EU leaders’ first AI summit, November 13, 2026
- The Next Web: Yann LeCun calls Anthropic’s Dario Amodei ‘deluded,’ Fortune reports, October 2, 2026
Researched with AI assistance, editorially reviewed by Robin Reuschel.
Robin Reuschel
Managing Director ROOVER
Robin Reuschel ist Experte für KI-Strategien bei Roover mit langjähriger Beratungserfahrung bei u.a. Roland Berger. Mit tiefer Expertise unterstützt er Unternehmen bei der KI-Transformation für zukünftige Geschäftsmodelle
You might also be interested in
AI transformation: How to implement AI correctly generated by Midjourney An AI transformation takes companies to the next level. It opens up new opportunities for increasing efficiency, automation and creates completely new business models. Companies are faced with the task of aligning their processes and structures with AI – a
Generative search is quietly but fundamentally changing the rules of the game. ChatGPT, Gemini and other systems no longer provide classic link lists, they respond directly. Often based on just a few sources classified as particularly trustworthy. The central question is therefore no longer: How well do I rank?but: Is