Anthropic just filed for a $1 trillion IPO.

Anthropic just filed for a $1 trillion IPO.

It has 300,000 enterprise clients.

Most of them also have an IT services contract with TCS, Infosys or Wipro.

Something has to give.

I was talking to a consulting leader last month who told me their biggest competitor isn’t another IT firm anymore. It’s the fact that their client’s product team had already solved the problem using Claude, before the SOW was even signed.

That’s the shift nobody in the Big Five wants to say out loud.

For years, the model was simple: enterprises had technology problems, IT firms supplied the people to solve them. Sticky contracts, long cycles, high switching costs. The relationship was the moat.

Anthropic didn’t attack that model. It just started solving problems before the model could kick in.

300,000 enterprise clients didn’t sign up for a vendor. They signed up because they had a problem, drafting, analysis, coding, a decision that needed a second opinion at 11pm and got an answer in seconds. No RFP. No kickoff call. No six-month implementation.

The IT services industry spent 3 years saying “AI is a tool we’ll use to serve clients better.”

Anthropic spent 3 years becoming the firm clients were already using.

Now they’re going public. Which means real capital, upmarket moves, enterprise SLAs, industry verticals, professional services wrappers. Everything the Big Five spent 30 years building, Anthropic can now buy or build in 36 months.

The firms that survive this won’t be the ones with the best AI stack. They’ll be the ones whose clients genuinely can’t imagine doing a transformation without them in the room. Not because of a contract. Because of earned trust and hard-won domain knowledge that no model has been trained on yet.

The question every IT services leader should be asking this week isn’t “how do we use Anthropic?”

It’s “what do we have that Anthropic can’t buy with IPO money?”