Google‘s chief executive has disclosed that roughly 75% of the company’s new code is now generated by artificial intelligence, a sharp rise from about 25% in 2024 and 50% in 2025, underscoring how rapidly automation is reshaping software development inside one of the world’s most influential technology firms; while engineers still review and approve the output, the shift signals a move toward so-called “agentic workflows,” where AI systems take on increasingly autonomous roles and dramatically accelerate engineering tasks—sometimes completing projects several times faster—placing human developers in more of an oversight and validation position rather than as primary creators of code.
Sources
https://www.semafor.com/article/04/24/2026/google-ceo-says-75-of-companys-new-code-is-ai-generated
https://www.reuters.com/business/google-puts-ai-agents-heart-its-enterprise-money-making-push-2026-04-22/
https://www.businessinsider.com/google-ai-generated-code-75-gemini-agents-software-2026-4
https://blog.google/innovation-and-ai/infrastructure-and-cloud/google-cloud/cloud-next-2026-sundar-pichai/
Key Takeaways
- AI has rapidly overtaken human-written code inside major tech firms, with Google reporting a jump from roughly one-quarter AI-generated code in 2024 to three-quarters today.
- Human engineers are increasingly acting as reviewers and orchestrators rather than primary coders, signaling a structural shift in the software workforce.
- The broader industry is following the same trajectory, with competitors aggressively expanding AI-driven development to boost productivity and reduce timelines.
In-Depth
What Google has revealed is not just a milestone—it is a flashing signal that the economics and culture of software development are being fundamentally rewritten. When three-quarters of new code inside a company of this scale is generated by machines, the traditional model of engineering—where skilled individuals painstakingly write, test, and refine code—gives way to something closer to industrial automation. Engineers are no longer the craftsmen; they are becoming supervisors of an automated production line.
There is an undeniable upside here. Productivity gains are real, and they are significant. Complex migrations and large-scale engineering tasks that once required months of coordinated effort can now be completed in a fraction of the time. That kind of acceleration does not just improve efficiency—it reshapes competitive dynamics. Firms that lean hardest into AI-driven development are likely to outpace those that hesitate, creating a widening gap between early adopters and laggards.
At the same time, the shift raises deeper questions that the industry has not fully answered. If machines are producing most of the code, accountability becomes murkier. Bugs, vulnerabilities, and systemic flaws may still trace back to human oversight, but the distance between creator and output is growing. That introduces new risks, especially in critical systems where reliability and security are non-negotiable.
There is also the workforce dimension. While it is fashionable to argue that engineers will simply “move up the stack” into higher-level roles, the reality is more complicated. Fewer people may be needed to produce the same—or greater—output. That does not eliminate the need for skilled professionals, but it does concentrate value in those who can manage, direct, and critically evaluate AI systems.
Ultimately, what Google is demonstrating is not just a technological evolution but a philosophical one. The act of coding is no longer the central value proposition; judgment is. The companies that succeed in this environment will not be the ones with the most engineers, but the ones that best understand how to harness machines without surrendering control to them.