Meta and Microsoft Say AI Will Soon Write Most of Their Code

Meta CEO Mark Zuckerberg and Microsoft CEO Satya Nadella have confirmed a transformative shift in how their companies write software: AI is set to take over a substantial share of coding responsibilities inside their companies. Speaking at Meta's LlamaCon 2025, Zuckerberg predicted that within a year, AI could be responsible for approximately 50% of Meta’s code—a rapid escalation in AI-assisted software engineering.

Satya Nadella speaking alongside him, shared that up to 30% of Microsoft’s code in select projects is already written by AI, largely through tools such as GitHub Copilot.

“I would say maybe 20–30% of the code that is inside of our repos today in some of our projects are probably all written by software,” Nadella said.

He emphasized that AI not only helps with code completion but is increasingly being used for code review, debugging, and automated task execution through agentic workflows.

Behind the Numbers: From Autocomplete to AI Agents

Zuckerberg acknowledged that while much of current AI coding is still in the autocomplete category, Meta is moving toward autonomous AI engineers capable of managing complex development tasks—particularly in support of the company’s Llama model ecosystem.

“Our bet is sort of that in the next year probably, maybe half the development is going to be done by AI as opposed to people, and then that will just kind of increase from there,” he said.

Microsoft, meanwhile, is deploying AI across greenfield and legacy codebases, with Nadella highlighting major strides in C#, Python, and increasingly C++—which had been less adaptable to AI assistance until recently.

Model Distillation and Developer Tools

Both companies are investing heavily in model distillation, a technique that compresses large AI systems into smaller, faster, and more efficient models—without sacrificing performance. This innovation makes high-powered AI tools more deployable, whether on enterprise servers or individual laptops.

At Microsoft, this is becoming central to the developer experience. Nadella described a future where engineers function as tech leads, overseeing fleets of AI agents. At Meta, Zuckerberg positioned these tools as essential to both internal development efficiency and broader open-source innovation.

What This Means

• For Software Developers: AI coding agents are poised to redefine developer roles, shifting the focus from manual code writing to reviewing, guiding, and collaborating with AI systems. While this raises concerns about job displacement, it also creates new leadership and integration opportunities.
  

• For the AI Industry: Meta and Microsoft’s announcements signal that AI is no longer just a product layer—it’s now a foundational element of engineering itself. Their investments may lower the barrier for smaller firms and independent developers to harness sophisticated AI capabilities.
  

• For the Future of Work: The pace of change suggests a historic shift in how software is built on par with the rise of cloud computing or mobile platforms. As generative AI becomes more capable, coding may evolve into curating and coordinating intelligent systems rather than crafting syntax by hand.

Looking Ahead

The message from both CEOs is unmistakable: AI is no longer an assistant—it’s becoming a partner in engineering. As development workflows transform, companies will need to redefine what it means to build software. The coders of tomorrow may be more like system architects, managing teams of intelligent agents.

The future of programming may not be about writing every line—but knowing how to direct the machines that will.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.