Twelve months ago, ChatGPT commanded roughly 87% of all AI chatbot web traffic. As of June 2026, that figure has fallen to 54.7%. Google Gemini has surged from 5.4% to 27.4% in the same period — growth of more than 100% in six months. Claude sits at 8.2% of consumer web visits, but wins roughly 70% of new enterprise deals when businesses choose between Anthropic and OpenAI.

These numbers — tracked monthly by First Page Sage and corroborated by Ramp’s AI spending index — tell the story of a market that has matured faster than almost any prior technology adoption wave. First-mover advantage erodes when competitors move at this speed.

What does this map look like for a business leader deciding which AI tools to invest in right now?

Why the Market Fractured So Fast

A year ago, the question “which AI should we use?” had an easy default answer: ChatGPT. It had the brand recognition, the developer ecosystem, the enterprise deals, and the head start. The choice was almost automatic.

That default is gone. Three forces drove the shift:

Google’s distribution play. Gemini’s growth is largely the story of Google’s scale. Gemini is now the default in the Gemini app, the engine behind Google AI Mode in Search, and following Apple’s WWDC 2026 announcement this week, the AI model powering a rebuilt Siri on roughly 1.4 billion iPhones by fall 2026. When a company can deploy their AI model through the world’s largest search engine and the world’s most premium mobile device simultaneously, the user base grows without product differentiation doing all the work.

Claude’s enterprise performance. Anthropic’s numbers look paradoxical on the surface: 8.2% of consumer web visits, but leading Ramp’s business adoption index and winning the majority of head-to-head deals against OpenAI among new enterprise customers. The reason is product-market fit. Claude’s reasoning performance, extended context window, and the breakout success of Claude Code with developer teams have made it the preferred infrastructure choice for businesses building serious workflows — not just experimenting with chatbot responses. Anthropic’s business adoption on the Ramp platform crossed 34.4% in April 2026, overtaking OpenAI at 32.3% for the first time.

OpenAI is defending from behind. ChatGPT crossed 1 billion monthly active users in June 2026, which is not a small number. But growth has slowed. OpenAI’s consumer dominance is real and durable for now, but the enterprise market — where budgets are larger and decisions are more deliberate — is no longer treating OpenAI as the automatic choice.

The Practitioner’s Dilemma

The fractured market creates a genuinely tricky decision for organisations that adopted one AI platform early. Sunk cost is real. Teams have built workflows, developers have learned APIs, and procurement has negotiated deals. Switching costs are non-trivial.

But the cost of staying on the wrong platform is also real. If your team is using ChatGPT for general queries but the developers building your internal tools would get meaningfully better output from Claude Code, or your analysts need longer context for the data work they’re actually doing — that gap compounds over time.

The pattern playing out in enterprise AI right now is not “one platform wins.” It is “different tools for different jobs,” with organisations running two or three AI platforms simultaneously: ChatGPT for general productivity, Claude for agentic coding and complex reasoning, Gemini increasingly embedded in Google Workspace workflows and now Apple devices.

What This Means for Business

Platform choice is now a real strategic decision. The defaults have shifted. Any organisation that last evaluated its AI tooling in 2024 is working off an outdated map.

Consumer share and enterprise fit are different metrics. Claude’s 8.2% web visit share versus its 70% enterprise win rate shows you cannot read enterprise suitability from consumer popularity. The question is not which AI is most popular — it is which AI performs best on your actual workloads.

The Apple-Gemini deal changes the stakes. When Gemini lands inside Siri on every modern iPhone later this year, every employee at your organisation will have a capable AI assistant in their pocket through the same device they already use for work. That is not a reason to adopt Gemini for your business infrastructure — it is a reason to get ahead of what your team will be asking about AI capability, and to make sure your own AI investments are at least as capable as what people are carrying in their pockets.

The multiplatform moment is here. The question is no longer “which AI platform should we pick.” It is “how do we build capability across whichever tools our teams actually use, and make sure we have the data foundations to get value from all of them.”

---

If your team is evaluating AI platforms or trying to get more value from your current investment, Enterprise DNA’s learning platform builds practical AI capability across your people — not just tool familiarity. For a more strategic look at where AI fits in your operations specifically, Omni Advisory can map that out with you directly.