Archived edition 14 November 2026 · Week 46

KI PowerUpdate

Your weekly AI briefing. From hype to need.

Go to the latest edition →

The week in two and a half minutes, read by a synthetic voice.

Duration: 2:15 min Download MP3 Podcast feed Read the script

01🧠 AI News of the Week

Handelsblatt

Nvidia without Jensen Huang: the chip giant faces a turning point

Jensen Huang has stepped down as Nvidia's chief executive, with Colette Kress leading the company on an interim basis. Huang shaped Nvidia into the engine of the AI boom, and his sudden exit could have consequences reaching into the US administration. Companies that depend on Nvidia hardware now have to ask how stable the supplier's strategy will remain.

Read the original article →

TechCrunch

Huang's departure leaves a hole in AI startup funding

Under Huang, Nvidia poured tens of billions of dollars into AI startups. Without him at the helm, founders and venture investors are bracing for smaller cheques and lower valuations. Young AI companies may need to reassess their reliance on a single strategic backer and broaden their sources of capital.

Read the original article →

t3n

SpaceX now earns a third of its revenue from Anthropic

A third of SpaceX's revenue now comes from the AI company Anthropic, and the US government is no longer its largest customer. The space company is effectively becoming an infrastructure provider for AI. The shift shows how strongly AI demand is reshaping business models far beyond the traditional software industry.

Read the original article →

TechCrunch

Google upgrades Search's AI Mode with Gemini 4 Hydrogen

Google's AI Mode in Search now runs on Gemini 4 Hydrogen and is meant to handle complex queries better, surface more relevant links and give richer visual answers. For businesses, this changes how content is found and presented. Visibility in search increasingly depends on how well content is prepared for AI-generated answers.

Read the original article →

02🚀 Tools & Startups

TechCrunch

LangChain reaches a $3 billion valuation

The startup LangChain says 7,000 customers use LangSmith, its model- and platform-agnostic tool for managing AI agents. The $3 billion valuation suggests that oversight and control of agents is becoming a market of its own. Companies deploying agents will increasingly need dedicated tooling to manage them.

Read the original article →

03⚖️ Regulation & Ethics

heise

Bundestag passes intelligence reform including AI surveillance

The German Bundestag has passed the contested reform: the BND and the domestic intelligence service gain far-reaching powers for home break-ins, AI surveillance and data manipulation. Warnings about violations of fundamental rights went unheeded. For companies and citizens, the balance between security and data protection shifts noticeably.

Read the original article →

04🎙️ Voices & Perspectives

05🏭 Business & Society

📄 Read the script

What does it mean for the AI industry when its most important hardware supplier suddenly loses its chief executive? Jensen Huang has resigned as head of Nvidia, and Colette Kress is running the company on an interim basis. Huang turned Nvidia into the engine of the AI boom, so his exit reaches far beyond one company.

The first mechanism is money. Under Huang, Nvidia put tens of billions of dollars into AI startups. Without him, founders and investors expect smaller rounds and lower valuations. A young company that counted on a single strategic backer now has to broaden its funding.

The second mechanism is politics. Democrats have opened a second congressional investigation into Nvidia's China sales and its dealings with President Trump. Alongside that, new research finds that China's most powerful AI models were trained on Nvidia chips under export bans. That questions how effective export controls really are, and it raises the risk of further tightening for anyone supplying internationally.

The customer side shows how resilient the industry still is. A third of SpaceX's revenue now comes from Anthropic, and the US government is no longer its largest customer. The space company is becoming an AI infrastructure provider. Anthropic itself is close to one of the biggest listings in history, with shares already six times oversubscribed. Amodei has to square growth expectations with a commitment to AI safety, and promising both at once is an exercise that rarely ends without compromise.

While the infrastructure wobbles and grows, the application layer is moving. Google now runs the AI Mode in Search on Gemini 4 Hydrogen and aims to answer complex questions better. For businesses, that means visibility increasingly depends on how well content is prepared for AI answers. ChatGPT Wallet goes a step further: AI agents shop and pay on users' behalf, starting with Pro subscribers in the United States.

Not everyone is going along with this. United Airlines blocks AI agents from booking flights, citing cybersecurity risks. Retailers therefore face a real decision: welcome agents as customers or keep them out. Both options carry costs, and neither can be waited out.

The public sector feels the pressure as well. In Brussels and Strasbourg, chatbots are producing a flood of amendments and enquiries, and officials warn of democratic risks. At the same time, the Bundestag has passed an intelligence reform giving the BND and domestic intelligence far-reaching powers for AI surveillance and data manipulation. The technology accelerates both the inputs to the state and the state's reach.

Put these threads together and the picture is consistent. The AI economy hangs on a few nodes, chips, capital and compute, and each node is more exposed to personalities and politics than the growth figures suggest. Anyone deploying or admitting agents needs not only tools but also a position on dependence and control. Stability in AI is currently less a question of technology than a question of leadership.

Sources in this edition: Handelsblatt, TechCrunch, heise, t3n