News Analysis
Google is sending a clear signal to other enterprises on the adoptability, effectiveness and scalability of AI-written code.
Sundar Pichai, CEO of Google and Alphabet, disclosed in a recent quarterly earnings call exactly how much Google is using AI to write its code:
"We're also using AI internally to improve our coding processes, which is boosting productivity and efficiency," Pichai said.
"Today, more than a quarter of all new code at Google is generated by AI, then reviewed and accepted by engineers. This helps our engineers do more and move faster."
His passing AI disclosure is easy to miss, but it's a major public marker in the AI and coding markets on AI moving from a coding capability and assistant for contributors to an operationalized enterprise reality at scale.
Here, we look at Google's announcement as a case study on the state of AI code in the enterprise, and some AI executives share how AI is being used for coding at their companies.
AI is Being Adopted to Write Enterprise-Wide Code
Pichai didn't qualify his statement, and it appears Google's use of AI extends well beyond an internal pilot to the widespread adoption of its own AI products — such as its Gemini family of foundation models, coding assistant and chatbot — for coding.
Internally, the adoption tells Google engineers Pichai and other company leaders believe in their AI products and their ability to enable software engineers to actually complete tasks and projects and deploy code.
Externally, the adoption tells the marketplace that Google is a user of the AI code products it's selling to enterprise development teams, and it has many first-hand learnings on AI code implementation to share in a young market.
Google's consulting and reseller partners will also benefit from this high-profile story as they seek client-side buy-in on large-scale AI code engagements.
Kamal Srinivasan, SVP of product at Parallels, a maker of cross-platform virtualization software, said that Parallels believes "AI is crucial for accelerating software development."
"Our approach is to use AI where it amplifies our productivity, helps reduce time to market and empowers our developers to deliver innovative solutions that directly benefit our users," Srinivasan said. "We see AI elevating and delivering on the promise of innovation and collaboration through paired programming."
Carlos Meléndez, co-founder and VP of operations at Wovenware, a software development company, said that "whether AI or human generated, coders always need to perform QA and review code."
"If anything, as AI get smarter, the review process could be reduced, since there will less errors and greater precision,” Meléndez said.
AI-Written Code Works Across the Enterprise
Google simply has too much to lose on many mission-critical fronts — financially, competitively, among its massive user base and in the public market — to deploy AI-written code that could be riddled with issues or doesn't work for end users.
After reportedly announcing an internal "code red" after OpenAI introduced ChatGPT, Google invested billions in AI and released a series of AI products, including for its key search, cloud computing, workspace and mobile divisions.
In the AI market, Google is facing fierce competition from several well-financed and fast-moving companies to lead the category, such as Microsoft, Meta, Amazon, OpenAI, Anthropic, xAI and Apple.
Google's billions of users across its integrated portfolio and billions of shareholders have one fundamental expectation of Google: for its software products to work well.
That said, then, the volumes of code written by AI for Google in the last quarter and reviewed by Google engineers are presumably effective and performing nearly as well as or better than code written by its developers.
At Parallels, for instance, Srinivasan said that with reasoning capabilities, large language models (LLMs) have significantly improved in code refactoring.
"These models intelligently consolidate redundant parts of code, providing clean, reusable code that maintains functionality," Srinivasan said.
AI-Written Code is Scalable in the Enterprise
Of the over 302,000 employees associated with Google on LinkedIn, over 78,600 work in engineering.
If Google used AI for 25% of new code company-wide in the last quarter, that could potentially be the equivalent of the previous collective output of over 19,600 engineers in a quarter.
Google is clearly replicating its AI coding process — such as its tools, frameworks and workflows — to what could be thousands of software engineers.
Significantly, the scale of the AI code adoption is presumably diverse: a mix of software engineers with varying code specialities working on multi-disciplinary projects and applications across its business units, departments and teams.
Srinivasan with Parallels said automated test generation for "fast-fail" development methodologies is a key area where AI in coding "stands out."
"We’ve found AI to be excellent at writing tests, saving us the time of manually crafting them," Srinivasan said. "This automated testing, generated alongside development, represents an ideal AI-test teammate that is transformative for the software development life cycle."
Webinar
Sep
25
CMS Briefing: A Live Look at What’s Next in AI-Driven Platforms
Learn how leading organizations are using AI‑driven tools to publish faster, personalize smarter and stay secure.
Register
Webinar
On demand
Ready or Not: How Data-First Organizations Are Unlocking Agentforce Potential
Learn how to cut through the noise, activate Agentforce and build a Salesforce AI strategy that actually delivers.
Watch Now
Webinar
On demand
AI in Customer Service: Faster Resolutions, Happier Customers
Don’t let rising demand burn out your team. See how to build a smarter, more resilient support org.
Watch Now
Webinar
On demand
From Hype to High-Impact CX Strategies That Actually Scale
Turn buzzworthy AI and outsourcing trends into measurable CX wins with fresh 2025 data.
Watch Now
Webinar
On demand
Insights to Action Rethinking the Contact Center for Real Business Impact
Join our exclusive webinar to hear CX executives share their innovative strategies for transforming service delivery.
Watch Now
Webinar
On demand
Accelerating Healthcare Ops with AI
Sign up to see how your peers are putting AI to work — and making it stick.
Watch Now
Expect Much More AI-Written Code at Enterprises and the Role of Coders to Change
Like all growth-oriented announcements in an earnings call, we can expect there to be future earnings calls when Pichai proclaims that a greater percentage of the company's code is created by AI: perhaps 50%, 75% and more in due time.
Of course, with this, the fundamental role of coders at Google is changing. If coders inherently code, a number of them are now spending much of their time reviewing AI-written code rather than coding.
Will there be a point where most of the company's software engineers are exclusively reviewing, testing and approving AI-written code? Will there be far fewer software engineers at Google, or will there be many more software engineers at Google to keep up the output of AI-generated code?
Srinivasan at Parallels said that at his company, "every developer is expected to be proficient in AI technologies," and it views "AI coding teammates with our human developer teammates."
Meléndez with Wovenware said “code generators are not replacing software developers — it’s just changing the way they work and the amount of code they can produce."
"AI is not fundamentally smart," Meléndez said. "Developers give it that intelligence, breaking down the coding problem into very small pieces.
"It’s important to remember that the key goal of AI-generated code is to augment the performance of human developers, not replace them."
