Anthropic CEO Dario Amodei Says Claude Now Writes 90% of Code, But Human Engineers Remain Indispensable

Artificial Intelligence continues to reshape the technological landscape, and its growing impact on software development is becoming increasingly evident. Anthropic CEO Dario Amodei recently stated that the company’s AI model, Claude, now writes around 90 percent of the firm’s code, yet human engineers remain vital to the process.

Speaking at the Dreamforce conference on Wednesday, Amodei highlighted how AI has reached a remarkable level of coding efficiency but clarified that humans are still at the center of software innovation and supervision.

“I made this prediction that, you know, in six months, 90% of code would be written by AI models. Some people think that prediction is wrong, but within Anthropic and within a number of companies that we work with, that is absolutely true now,” Amodei said, as quoted by a popular publication.

Amodei’s revelation signals a major transformation in how companies approach software development. The use of AI in coding, sometimes referred to as “vibe coding,” is expanding rapidly across the tech sector. Anthropic’s latest model, Claude Sonnet 4.5, reportedly has the ability to generate up to 11,000 lines of code independently, showcasing the rapid evolution of generative AI in programming.

While the rise of AI-driven coding tools has sparked fears of large-scale job losses in the tech industry—especially after layoffs at major firms such as TCS and Accenture—Amodei took a more balanced stance. He stressed that AI should not be viewed as a replacement for human coders but rather as a tool that enhances their capabilities.

“You might need more, because they can then be more leverage,” he explained, suggesting that developers could become ten times more productive when working alongside AI tools.

According to Amodei, the collaboration between humans and AI represents a new phase in software engineering—one where AI handles most of the repetitive coding tasks, and engineers focus on creative problem-solving, system design, and quality supervision.

“They can focus on the 10% that's editing the code or writing the 10% that's the hardest, or supervising a group of AI models. And so, what happens is, you know, you just end up being 10 times more productive,” he elaborated.

Despite his optimism about AI-augmented productivity, Amodei has previously acknowledged the potential challenges ahead. Earlier this year, he warned that AI could eliminate about 50 percent of entry-level tech jobs within the next five years, raising questions about how the workforce will adapt to this shift.

Still, the CEO’s latest remarks underline a key message: while AI like Claude is revolutionising software creation, human engineers remain irreplaceable. Their expertise, intuition, and creativity continue to be the backbone of innovation—even in an era where machines can write most of the code.