Key Takeaways
- A new report from Google finds that 90% of software developers are using AI at work, with 65% confirming that they are heavily reliant on the technology.
- Writing new code is the most popular use case, as confirmed by 71% of respondents. Next up is modifying existing code (66%).
- Report also cautions against blindly adopting AI, with fragmented organizations likely to gain minimal benefits.
A staggering 90% of software developers now use AI to carry out their day-to-day tasks, a new report from Google reveals. The State of AI-assisted Software Development surveyed nearly 5,000 technology professionals around the world to better understand how software is built, finding that AI is now ubiquitous among developers.
This figure represents a massive 14% increase from last year, with these employees dedicating a median of two hours per day to working with different AI tools. The most popular task is writing new code, with 71% of respondents confirming that that is what they commonly use AI for.
The report adds to a growing body of evidence that, in recent years, AI has revolutionized the way that we carry out our work. However, it also cautions against blindly pursuing an AI strategy without first making significantly organizational and cultural changes to accommodate the shift.
90% of Software Developers Now Using AI at Work
The overwhelming majority of software developers now use AI as a key part of their everyday operations, according to new research from Google. The 2025 DORA report, titled State of AI-assisted Software Development, finds that no fewer than 90% of developers confirmed that they now use the technology at work.
Nearly 5,000 respondents took part in the report, which takes a yearly look at the “capabilities and conditions that fuel the success” of tech companies around the world, according to the website. Perhaps unsurprisingly, the latest edition focused on how AI was changing the way that technology teams work.
This just in! View
the top business tech deals for 2025 👨💻
As per the study, most participants (65%) confirmed that they are heavily reliant on AI for software development, with a further 37% reporting a “moderate amount” of reliance.
Impact of AI Largely Positive, But Questions Remain
Other findings shed light on how software developers are leveraging AI in their day-to-day. Most (71%) are using it to write new code, while an additional 66% are modifying existing code with AI. After that, 64% are writing documentation, and 62% are creating test cases and explaining concepts.
No matter the use case, the general consensus is that AI is an unqualified success. For example, over 80% of respondents have indicated that AI has enhanced their productivity, while 59% noted the positive influence of AI on code quality.
Despite this, the Google report also observed that, on an organizational level, the impact of AI is “more complex.” While it is enabling businesses to ship more software and applications, there remains the issue of quality control.
“Think Before You Act”
More broadly, Google researchers observed that AI can act as both a “mirror and a multiplier.” In other words, organizations that are well-established to accommodate the technology will thrive, while fragmented businesses will find their weaknesses reflected back at them.
As our own research here at Tech.co demonstrates, businesses don’t always pause before pulling the trigger on their AI strategies. In fact, 58% of surveyed respondents confirmed that “pressure from competitors” was the single biggest driver for adopting AI. This “act first, ask questions later” approach does not guarantee success.
Setting out a “blueprint for guiding AI in organizations,” Google claimed that it’s essential for technology businesses to focus on seven key areas alongside taking up the technology, including team performance, code quality, and more. If followed, your AI implementation should start to bear fruit.
