The short version: ChatGPT's share of the AI tool market has fallen from 76.5% in February 2025 to 54.7% in June 2026. Google Gemini has climbed to 27.4%. Claude sits at 8.2% globally — but with 306% quarterly growth, the fastest acceleration of any major model. The market is maturing, competition is real, and "I use ChatGPT" is no longer a complete answer to the question of which AI to use for business.

The data: what the market shift looks like

In February 2025, ChatGPT held 76.5% of the AI tool market. Eighteen months later, that has fallen to 54.7% — still the majority, but a dramatically reduced one. Google Gemini has risen from under 10% to 27.4%, driven largely by its integration into Google Workspace and Android. Claude has reached 8.2% globally, with 306% growth in a single quarter.

The pattern is consistent with how most technology markets mature: a dominant first mover holds near-monopoly share in the early years, then competition arrives, quality differentiates, and the market consolidates around two or three credible options. Email, social media, and cloud storage all followed the same curve. AI tool adoption appears to be following it faster.

Why market maturity matters to UK small businesses

When ChatGPT held 76% of the market and was clearly the default choice, the answer to "which AI should I use?" was simple. The market maturity now means that question requires a different kind of answer — one based on capability fit rather than default choice.

Three distinct use cases are emerging with different tool advantages:

  • Conversational tasks, ChatGPT: Quick answers, general research, chat-style interactions. Still the most accessible entry point for AI-curious users.
  • Workspace-integrated AI, Gemini: Businesses fully embedded in Google Workspace (Gmail, Drive, Docs) benefit from Gemini's native integration. The 27.4% share reflects how many UK SMBs are on Google tools.
  • Complex document and workflow tasks, Claude: Longer documents, precise instruction following, multi-step reasoning, and the highest benchmark scores. The 306% growth reflects developers and businesses discovering this performance gap through use, not marketing.

The practical implication for AI tool choices

If your business is making an AI platform choice in the next few weeks — whether choosing your first tool or reviewing existing usage — the market data suggests a two-step approach: first, map your most common use cases (email drafting, document processing, customer communication, data analysis); then test the two or three tools that suit those use cases rather than defaulting to the highest-profile name.

Many UK businesses that tried AI in 2023 or 2024 and found it inconsistent or unreliable tried tools at a point when the market was far less mature. The models available now — including the latest Claude and Gemini versions — are measurably more capable. A re-evaluation in 2026 may produce a very different result.

What Claude's growth signal says specifically

306% quarterly growth is not sustainable indefinitely, but it indicates something specific: businesses that try Claude in the context of real work tasks are finding it better than their current tool. Growth at that rate is almost entirely word-of-mouth and direct discovery, not advertising. For a UK service business owner considering a switch or a first proper AI tool investment, that is a meaningful signal.

Operator move for this week

List the three AI tasks you attempt most regularly. Run each one through the free tier of Claude (claude.ai) and compare the output to what your current tool produces. Quality differences that are invisible in demos become obvious on real business tasks. Thirty minutes of testing beats any benchmarking article.