France: OpenAI, Google, and Anthropic control over 84% of the AI agent market

France's antitrust regulator, the Autorité de la concurrence, published an opinion on July 17, 2026 (Avis 26-A-05) — more than 3,700 pages including annexes — concluding that OpenAI, Google, and Anthropic together control more than 84% of the global AI agent market, ppc.land reports.

What worries the regulator

• "Platformisation" — risk of discrimination and self-preferencing by vertically integrated players.
• Disintermediation — the shift from chatbots to autonomous agents could accelerate concentration of the digital economy around a small number of firms.
• Algorithmic collusion risk in the agentic-commerce sector.
• User lock-in effects within a specific ecosystem.

What the regulator recommends

According to the official press release, the Autorité de la concurrence calls for interoperability and data portability, and for new standards to be

“developed and maintained in a transparent, open and collaborative manner, to avoid excessive control by dominant player”

— from the Autorité de la concurrence's official opinion

Quote source: ppc.land.

Scale and context

The inquiry follows the regulator's earlier reviews of cloud computing (2023) and generative AI (2024), plus a December study on AI's energy impact. AI-agent traffic to e-commerce sites currently sits below 5%, but the authority estimates it could grow to 20-25% by 2030. Other market players named include Amazon, Microsoft, Nvidia, Mistral AI, Perplexity AI, and xAI — but the authority says it's the top three that are setting the rules for the whole sector. The theme of power concentrating around a handful of AI companies echoes the technology-theft accusations we covered yesterday in the White House-Moonshot AI dispute.

None of this should be read as personalized investment advice.