Google says AI now writes 75% of all codes– Are human developers becoming irrelevant?

When it comes to coding, most of the tech companies are turning to artificial intelligence to ensure accuracy and efficiency.

After Anthropic, Google is the latest company at which 75 percent of new computer codes are generated by AI models and approved by engineers, a revelation made by Google’s CEO Sundar Pichai while speaking at the Google Cloud Next 2026 conference in Las Vegas.

The recent statistics reveal a stark yet sudden surge in the use of AI models for coding purposes. In 2024, the usage was 25 percent; it reached 50 percent in 2025.

“We’ve been using AI to generate code internally at Google for a while,” Pichai said.

“We are now shifting to truly agentic workflows. Our engineers are orchestrating fully autonomous digital task forces, firing off agents, and accomplishing incredible things,” Google’s chief executive added.

According to the Google boss, the purpose of using artificial intelligence for coding is to boost efficiency , as AI agents are capable of completing complex code migration tasks 6-times faster than human engineers.

In recent times, most tech companies are integrating AI into their workflows on an unprecedented scale. Earlier this week, Meta announced the plans to train AI models by tracking employees’ clicks, keystrokes, and data activities.

A new tool called the Model Capability Initiative (MCI), unveiled by Meta, aims to replicate or evaluate how humans interact with computers.

Similarly, OpenAI is also reportedly ranking staffers on the ground of their monthly token consumption, called tokenmaxxing.

This approach shows the value of AI power usage values over real performance metrics.

Nvidia CEO Jensen Huang is pushing for a future where his engineers spend zero percent of their time manually writing code.

During a recent appearance on the No Priors AI podcast, Huang shared that every engineer at the $3 trillion chip giant has integrated Cursor, an AI-powered coding assistant, into their daily workflow. His objective is to liberate his team from the mechanical "syntax" of programming.

"Nothing would give me more joy than if none of our engineers were coding at all. And they were just purely solving undiscovered problems, Huang said.

Dario Amodei, the CEO of Anthropic, also vouched for using AI for the coding tasks.

"I think we will be there in three to six months, where AI is writing 90% of the code. And then, in 12 months, we may be in a world where AI is writing essentially all of the code," Amodei said last year in an event.

Given the heavy integration of AI in coding tasks, it remains to be seen whether human developers will be at helm or lose their relevancy in the age of AI.