← Latest AI News

October 28, 2026 · Daily brief

Anthropic made $16 billion in six months and turned a profit

Sovereignty angle
The loudest knock on AI was that the business wouldn't hold up. Anthropic just answered: $16 billion in six months, a profitable quarter, and nearly all the money it raised still in the bank. The debate moves from whether it works to what it costs you.

Anthropic filed to go public, showing $16.3 billion revenue in the first half of 2026 and a $974 million operating profit in Q2—proof an AI business can work at scale.

Anthropic filed its S-1 with the SEC on October 26, targeting a Nasdaq listing under ticker ANT. The documents show revenue hit $16.3 billion in the first six months of 2026, about 14 times the year before. The June quarter alone brought $11.6 billion and turned a $974 million operating profit.

Anthropic has raised $141 billion to date but still holds about $132 billion. That means it has spent a net of roughly $9.4 billion to build a business making $11.6 billion a quarter. The June operating profit proves that an AI company can make more than it spends on servers and people.

What you can do with it now

The filing proves the tools you're considering have revenue behind them. Claude is part of a business making real money, not a lab demo. That means you can expect the service to still be there in twelve months when you've trained your team on it.

If you've been piloting Claude or ChatGPT with one person, this is your signal to roll it out to three more. Pick the Friday task everyone avoids: summarizing the week's customer questions, drafting follow-up emails, pulling numbers from five different reports. Give the tool to the person who owns that task and check the output together every Friday for a month.

What to keep in your own hands

Don't paste your full customer list, your pricing, or your supplier terms into any AI tool. Use it for the draft, the summary, the first version—but keep the decisions and the sensitive data with a person. And keep a second option ready: if your team depends on one tool and the price doubles, you need to know you can switch in a week, not rebuild for six months.