How AI generates 75 percent of Google's new code

Introduction

The technology industry is undergoing one of the most rapid transformations in the last 50 years, and the integration of artificial intelligence into software development processes is the central driver of this evolution. The recent statement by the Google CEO, according to which approximately 75% of new code written internally is generated by AI, marks a defining moment for the entire industry DevOpsThis shift is not just an operational optimization, but a new collaboration model between developers and advanced AI systems, which is changing the way organizations approach the development, testing, and delivery cycles of modern applications.

By adopting AI code generation tools, Google confirms what many companies are beginning to discover: the productivity, quality, and speed at which software is built can be drastically improved when AI is implemented methodically and strategically. At the same time, this shift opens up new debates about the future of the role of programmers, accountability for automatically generated code, and the reconfiguration of processes. DevOps in a world where automation is reaching unprecedented heights.

The rise of AI as a software production engine

Modern code generators, like Gemini, Codey or similar platforms, are no longer simple autocomplete tools. They understand context, analyze complex repos, assimilate programming styles and can deliver functional, testable and continuous integration-ready code. Where in the past AI could only suggest simple snippets, now it can build modules, APIs, infrastructure as code and complete CI/CD pipelines.

At Google, the widespread adoption of AI systems has led to significant efficiency gains across all stages of the SDLC. AI tools can:
• analyze product requirements and generate initial code with high accuracy; • propose architectural redesigns based on modern best practices; • detect vulnerabilities and recommend fixes before code review; • automatically produce unit tests, integration tests and technical documentation.

These capabilities radically transform the way teams work, allowing developers to focus on conceptual design, architecture, and logical validation of solutions, while AI handles repetitive or time-consuming tasks.

Why is 75% such an important threshold?

Reaching the 75% mark of automatically generated code is, without exaggeration, a historic inflection point. It reflects not only the maturity of the technology, but also the trust that Google places in these systems. In fact, three-quarters of the new code is no longer written manually, which shows that AI is no longer an optional assistant, but an essential component of the pipeline DevOps.

For the industry, this figure sends a clear message: organizations that do not adopt similar tools will fall behind in terms of development speed, scalability and competitiveness. Moreover, the 75% threshold suggests that, in the near future, the role of AI will become even more extensive, perhaps even exceeding current responsibilities by completely generating end-to-end systems, with minimal human intervention.

The impact of automatic code generation on teams DevOps

DevOps is a defined area of ​​automation, rapid collaboration, and continuous integration, and AI is perfectly aligned with these principles. With the right tools, AI can function as a virtual team member, providing ongoing support and reducing intense operational tasks. For professionals DevOps, this equates to more time for architecture optimization, advanced security, observability, and increased application performance.

Thus, AI directly contributes to the maturation of complex pipelines, and brings improvements in areas such as:
• automation of continuous integration and delivery; • infrastructure configuration cloud at scale;• incident anticipation and automatic remediation;• observability analysis and rapid generation of post-mortem reports;• validation of compliance and security policies.

This acceleration does not compromise quality, but rather improves it, as AI can detect subtle errors that manual teams often miss, especially on large and complex projects. In addition, AI tools are trained to respect internal code patterns, contributing to a consistent and easy-to-maintain style.

Generating infrastructure as code with AI support

One of the most impressive current applications of AI in the ecosystem DevOps is to assist in the generation of infrastructure as code. Using high-level instructions, teams can define architectures cloud complex, and the AI ​​takes care of creating Terraform scripts, Helm Charts, Kubernetes pipelines, and multi-configurationscloudThus, configurations that normally required days of work can be produced in minutes.

This type of automation reduces the risk of misconfigurations, optimizes costs, and allows rapid testing of different architectural scenarios, accelerating teams' experimentation and innovation strategy.

What does this evolution mean for programmers?

An essential aspect to clarify is that AI does not replace programmers, but radically changes their role. Programmers are no longer code executors, but orchestrators of an intelligent ecosystem. They become architects, designers, analysts and auditors of automatically generated code. In other words, the value brought by specialists increases, and the nature of the work is oriented towards logic, strategy and creativity.

AI also removes barriers for those just starting out, allowing juniors to produce advanced results in record time without compromising quality. In this way, AI democratizes access to complex software development.

AI and software security

Security remains one of the most sensitive topics in AI adoption. Automatically generated code must be carefully vetted to avoid introducing vulnerabilities. However, when configured correctly, AI can bring enormous benefits to application security:
• high-speed static and dynamic analysis; • identification of risk patterns before implementation; • automatic implementation of zero-trust policies; • best practice recommendations based on extensive data sets.

Google is a significant example due to its huge volume of data and high security standards, demonstrating that AI can manage code at scale without compromising the security of the ecosystem.

What's next for the industry DevOps

The adoption of AI for code generation is just the beginning. In the coming years we will see:
• AIs capable of managing distributed applications completely autonomously;• systems that continuously learn the internal models of companies;• 100% autonomous CI/CD pipelines, with human interventions only at the decision-making level;• tools that design architecture based on budget, SLA and KPI;• DevOps intelligent, based on anticipatory observability and self-healing.

Google has led the way, but the entire tech ecosystem will follow a similar path. The companies that adapt will dominate the next digital era.

Conclusion

The fact that Google generates 75% of its code with the help of AI is a clear sign of the maturity of new technologies. The impact on DevOps, on the way developers work and on modern software architectures is enormous and growing. As AI becomes an indispensable partner in development processes, the future of coding will be defined by the close collaboration between humans and intelligent systems. And the companies that know how to harness this force will dictate the pace of innovation in the next decade.

Surely you understood what the news in 2026 is related to DevOpsIf you are interested in deepening your knowledge in the field, we invite you to explore our range of courses structured by roles and categories in DevOps HUBWhether you're just starting out or want to brush up on your skills, we have a course for you.