Selling Intelligence Was the Warm-Up For AI. Owning Vertical Markets May Be The Play

As enterprise artificial intelligence evolves, selling intelligence appears to be losing its allure in the boardroom. Instead, the focus is shifting towards selling a model to a company, integrating it into their most valuable work, learning the business from the inside, and then competing against it with a product of their own.

This pattern has become evident in recent months, with frontier labs moving into their customers' markets openly. The strategy involves selling the model to the world's most valuable businesses, helping wire it into their sensitive and highest-margin work, mapping how the business runs from the inside, and then turning those learnings into a product to compete with the customer directly.

While execution remains crucial, this approach allows the labs to convert supplier relationships into market entrants more effectively. Figma's partnership with Anthropic serves as a prime example, as the tech company launched its own AI design tool, Claude Design, after Anthropic's chief product officer resigned from Figma's board.

Similarly, Harvey, a legal startup, transitioned from being a customer success story to a competitor for Anthropic, offering similar workflows for law firms. This playbook extends to various industries, including financial services, where Anthropic's Claude is used for market data providers and credit memos, and pharma, where the model is employed in drug discovery programs.

Despite the potential risks, the economic incentives make this strategy difficult to resist, as companies with large valuations and public listings cannot sustainably leave the richest layers of the stack to their customers forever.