It was one of those weeks where the tech news felt like it was coming at you from every direction at once. The founder-CEO of the world’s most valuable chipmaker resigned in disgrace, the AI lab behind Claude put a trillion-dollar price tag on its IPO, Google quietly swapped the brain inside Search, and Microsoft shipped another four-figure patch haul. Here’s the week that was.

Jensen Huang resigns from Nvidia

The biggest story of the week, and not a happy one. Jensen Huang — the co-founder who ran Nvidia for 33 years and turned a graphics-chip company into the beating heart of the AI boom — resigned as president, CEO, and board member on Thursday, November 12, effective immediately. Nvidia’s own announcement named CFO Colette Kress as interim CEO while the board searches for a permanent successor, and appointed director Harvey C. Jones as the company’s first-ever board chair.

The resignation capped a brutal week. On Sunday, November 8, Nvidia’s board announced an independent investigation into whether Huang had been targeted by a foreign intelligence operation, after reports of his relationship with a Singapore-based tech executive whom U.S. officials suspect of spying for China. Huang admitted to an extramarital affair that Sunday and apologized to his family and colleagues. Then the New York Times reported a second affair, this one with a Nvidia senior executive, and the Wall Street Journal reported the board was preparing to oust him. The board had initially said it had confidence in Huang; by Wednesday night it had changed its mind.

The market noticed: Nvidia shares slid more than 10% in pre-market trading Thursday and closed down 4.3% at $166.19, their lowest close since March 30, according to Barron’s. There is no obvious successor — Huang had no succession plan, with dozens of executives reporting directly to him. The next test comes fast: Nvidia reports third-quarter results on Tuesday, November 17, as previously scheduled, and Kress will be the one on the earnings call.

Anthropic’s IPO gets a price: $70 to $78 a share

Speaking of the AI boom getting a price tag — Anthropic, the company behind Claude, set the terms for what could be the largest IPO in history. In an updated SEC filing on November 10, the company said it plans to sell 825 million Class A shares at $70 to $78 each, pricing on Wednesday, November 18, with trading expected to begin on the Nasdaq under the ticker ANT on Thursday, November 19. At the top of the range, the fully diluted valuation reaches roughly $1.4 trillion. TechCrunch’s reading of the filing puts the public raise at $57.75–$64.35 billion — and that’s before a separate $26 billion private placement of non-voting stock to seven companies: Google, Nvidia, AMD, Broadcom, Micron, Microsoft, and TSMC.

Demand looks fierce: reports put the order book at roughly six times oversubscribed, and bankers are said to be exploring pricing above the range. The amended filing also disclosed run-rate revenue of $81 billion at the end of October, up $10 billion in a single month, according to Benchmark’s read of the filing. For a company whose last private round priced it around $60 billion, the step up is staggering — and Wednesday’s pricing will tell us whether public markets agree.

Gemini 4 Hydrogen is now the default brain of Google Search

Google rolled out Gemini 4 Hydrogen on November 12, and unlike the flagship Carbon model that arrived for Ultra subscribers two days earlier, this one is for everyone: Hydrogen is now the default model in Google’s AI Mode globally, and it handles the hardest questions in AI Overviews too. Google calls it the biggest model upgrade to Search yet.

Hydrogen is the fast, efficient member of the Gemini 4 family — Google says it completes tasks 2.5 times faster than Gemini 3.8 Flash, the model it replaces as the default. The practical pitch: better handling of long, multi-part questions (even ones full of typos), improved “query fan-out” that runs many related searches in parallel to surface a broader range of sources, and more visual answers with comparison tables and images. It’s also live in the Gemini app, Google AI Studio, the Gemini API, and Antigravity for Pro and Ultra subscribers. If your searches started feeling a little sharper this week, this is why.

Patch Tuesday: 1,235 fixes and an Outlook zero-day already being exploited

Microsoft’s November Patch Tuesday, released November 10, fixed 1,235 vulnerabilities — the second month in a row over 1,000, after October’s record 1,411. Sixty-six of them are rated Critical. But the number isn’t the story; the one actively exploited flaw is. CVE-2026-124194 is a spoofing vulnerability in classic Outlook: a specially crafted message file can make Outlook silently authenticate to an attacker-controlled server, leaking the user’s NTLMv2 password hash. Microsoft rates it Moderate with a CVSS score of 6.5, but CISA added it to its Known Exploited Vulnerabilities catalog the same day patches shipped, giving U.S. federal agencies until November 24 to fix it.

Two things worth knowing if you run Outlook anywhere: the fix ships through Office updates, not just Windows Update, so patch pipelines that stop at the OS level will miss it — and the attack requires opening the crafted file, not just previewing it in the reading pane. Patch Outlook, then the rest.

Claude moves into the office

Anthropic had a busy week beyond the IPO roadshow. On November 10 it launched a bundle of workplace tools that push Claude from coding assistant into the meeting room: Claude Sheets and Claude Whiteboards are in beta on all paid plans — shared spreadsheets with tabs and formulas that Claude builds and edits alongside you, and a shared whiteboard where Claude organizes brainstorming into timelines, owners, and deadlines, then converts the board into a doc or slide deck. Claude Data (formerly Dashboards) is now generally available.

The headline act is Claude Meetings, which joins Google Meet calls as a visible participant — announcing itself in the chat, answering questions during the discussion, and producing notes, plans, and action items afterward. It’s in beta for Team and Enterprise plans, though Anthropic’s own announcement still labels it “coming soon,” so don’t be surprised if it isn’t switched on for your workspace yet. Either way, the direction of travel is clear: every AI lab now wants to be where your work actually happens, not just where you ask it questions.