The Best AI Doesn’t Always Win — Anthropic Just Proved It

Anthropic has overtaken OpenAI in enterprise AI spending in the United States, now leading with an impressive 70% market share — despite OpenAI holding roughly 90% of the market just one year ago.

This reversal is happening in spite of ChatGPT’s historical head start and its massive user base of approximately 900 million weekly users. So what exactly changed this year? The simple answer is that this was never a battle of models — of who is smarter or more accurate. It was a procurement battle.

Consumers choose AI tools based on capability and user experience. Organizations, on the other hand, choose based on risk, regulation, data security, and contractual convenience. Once AI budgets shift from the experimentation phase to a permanent budget line item, the decision-makers change. Legal counsel, security managers, and finance directors enter the picture — and for them, the most powerful model does not necessarily win. The winner is the vendor that earns the greatest trust.

Anthropic recognized this early. From the outset, it focused its efforts on building solutions tailored to enterprise procurement and reducing legal friction — and this created a decisive competitive advantage precisely at the moment the market matured: when budgets moved from early pilots to broad deployment. This story is an important reminder for anyone in the field — in a business market, being the best is not enough. You need to be the most trusted.