🤑 Anthropic closes $30B, sues the Pentagon, and Gemini Spark goes live

$30 billion in a single round. $900 billion valuation. A startup spending $1.25 billion a month on compute alone, close to profitability. Anthropic just closed the largest funding round in history. And in the same week, it sues the Pentagon for not wanting to build autonomous weapons, and Google launches the first real 24/7 agent for consumers. The company that raised the most money is also the only one saying no to the US government. Such are the contradictions of AI in 2026.

🤑 Anthropic closes $30B at $900B: the world's most valuable startup

The funding round that financial journalists had been tracking for two weeks officially closed between Monday and Wednesday of this week. Anthropic has raised more than $30 billion at a pre-money valuation exceeding $900 billion, surpassing OpenAI's $852 billion private valuation from March. This in the context of the triple IPO of SpaceX, OpenAI, and Anthropic.

The round was co-led by four firms: Sequoia Capital, Dragoneer Investment Group, Altimeter Capital, and Greenoaks Capital Partners, each committing approximately $2 billion. Peter Thiel's Founders Fund and General Catalyst also participated. The round closed in less than four weeks from first investor contact to close — Bloomberg describes it as unusually fast for a round of this size.

The numbers behind the valuation are equally staggering. Anthropic projects $10.9 billion in revenue in Q2 2026, 130% more than the $4.8B in Q1, alongside its first quarterly operating profit of approximately $559 million. Compute costs are falling from 71 cents per dollar of revenue in Q1 to a projected 56 cents in Q2. The annualized revenue run-rate will exceed $50 billion by end of June.

The number nobody talks about: Anthropic spends $1.25 billion a month on SpaceX compute alone — a figure that only became public when SpaceX filed its IPO prospectus. Add AWS, Google Cloud, and other providers, and the infrastructure spend is extraordinary. That the company is close to profitability despite that says something extraordinary about how fast revenue is growing.

From the numbers to the courage: the same week it closes the largest round in history, Anthropic stands up to the US government.

⚖️ Anthropic sues the Pentagon: the red line on autonomous weapons

And now the other side. The same week it closes the largest round in history, Anthropic filed a lawsuit against the US Department of Defense.

In March 2026, the DoD designated Anthropic as a "supply chain risk" after the company refused to allow its technology to be used for lethal autonomous weapon systems or mass surveillance of US citizens. The designation meant that companies doing business with the Pentagon would face restrictions on using Claude's models.

A federal judge issued a preliminary injunction blocking the designation while the case is litigated. Anthropic estimates the dispute puts hundreds of millions to billions of dollars of 2026 revenue at risk.

The key point: Anthropic is the only frontier lab that has drawn a red line and is willing to litigate to defend it. OpenAI signed a partnership with the DoD for non-lethal military use. Microsoft is a major defense contractor. Google reversed its position on AI in drones after Project Maven. Anthropic says "no" and goes to court.

Pope Leo XIV's encyclical, published the same week, directly addresses autonomous weapons. Christopher Olah presenting at the Vatican while Anthropic litigates against the Pentagon over exactly this issue is not casual positioning — it's a coherent strategic statement about what kind of AI company Anthropic wants to be.

And while Anthropic litigates, Google shows consumer AI is also advancing — with the first agent that works while you sleep.

🤖 Gemini Spark is here: the first real 24/7 agent for consumers

Google launched Gemini Spark this week for Google AI Ultra subscribers in the US at $100/month. It's the most ambitious consumer AI product Google has launched in years, and the first real 24/7 agent a consumer can activate today.

Spark runs on Google Cloud virtual machines even when your laptop is closed. You give it a task and it works in the background across Gmail, Google Sheets, Google Docs, Google Drive, and Calendar, showing updates through Android Halo — a new notification layer in the Android status bar. Every action requires user approval before execution.

The launch demo at Google I/O showed Spark planning a block party: pulling RSVPs from Gmail, tracking who was bringing what, contacting those who hadn't replied, building a live tracker in Sheets, and generating a Slides presentation with details about the bouncy castle and local rules extracted from a Drive file. All without the user opening any app.

MCP for third-party apps like Canva, Instacart, and OpenTable arrives in weeks. Chrome integration arrives this summer.

OpenAI has Operator, but it's still limited. Anthropic's agent platform is powerful but enterprise-focused. If Spark delivers what the demo promised at the Ultra price point, Google wins the consumer agent category in 2026.

Anthropic closes $30B, sues the Pentagon, and Google launches the first real 24/7 agent. One week where capital, ethics, and product converge. The company that raised the most money in history is also the only one saying no to the Pentagon. Google, meanwhile, puts a real agent in consumer hands. The message is clear: building AI at scale while maintaining principles isn't incompatible — it's the only sustainable way to do it.

— Max