Anthropic’s Amodei says AI coding market is maturing

Dario Amodei has a habit of saying quiet parts loudly. The Anthropic CEO’s latest signal to the market: the AI coding boom is no longer a boom. It is becoming infrastructure.

Amodei has indicated that the enterprise AI coding market is now valued at approximately $34 billion annually, with Anthropic’s Claude Code and OpenAI’s Codex together capturing roughly 70% of that total, or about $24 billion combined. That is not a niche. That is a duopoly forming in real time.

The numbers inside Anthropic’s own walls

The most striking data points are not market-wide. They are internal.

As of September 2026, AI models in the Claude family were responsible for over 90% of code contributions at Anthropic itself. That figure alone reframes what it means to call a company an “AI lab” versus a company that simply employs software engineers.

The productivity shift is just as dramatic when measured per person. By the second quarter of 2026, the average lines of code merged per contributor had reached eight times the pre-2025 baseline.

Amodei has also said that Anthropic’s R&D pace has doubled year-over-year, a trend the company attributes directly to its AI tooling.

The workforce implications are real and already visible. As AI absorbs routine coding tasks, demand for entry-level technical positions at Anthropic has declined. Amodei has acknowledged that certain functions, specifically product design and team management, still require human judgment.

A timeline that keeps moving forward

Amodei suggested earlier in 2026 that AI could potentially handle nearly all software engineering tasks within six to twelve months from that point. That is a narrower window than most of his peers in the industry have been willing to put on record.

The qualifier matters: “potentially.” Amodei is not claiming autonomous software engineering is already here. He is claiming it is close enough to plan around.

The broader competitive picture includes OpenAI, whose Codex product shares market leadership with Claude Code, and Google, which is pushing hard into the same territory with its own developer-facing AI products. The fact that two companies already control roughly 70% of a $34 billion market suggests consolidation is happening faster than most industry analysts predicted even twelve months ago.

What Anthropic’s IPO plans add to the story

Anthropic is reportedly targeting a potential IPO by November 2026. The timing is not incidental. The company would be going public during what appears to be the peak commercial moment for AI coding tools, when revenue is growing, the market narrative is intact, and the internal productivity metrics give public investors a concrete story to price.

Amodei’s public comments about AI safety remain central to Anthropic’s brand identity, particularly as the company approaches a moment where its products are reshaping employment patterns in a visible way. The company has consistently positioned safety and capability advancement as complementary rather than competing priorities.