new york
–
The vast majority of high-tech industry workers use artificial intelligence at work for tasks like writing code and modifying, a new Google study found.
A report from Google's Dora Research Division found that 90% of respondents use AI from 14% last year, based on 5,000 responses from technology experts around the world.
The rise of AI has sparked a mix of concerns, hype and promises about how technology will affect employment and the economy, resulting in research findings. Humanity CEO Dario Amody made the headline in May when he said that AI could lead to a surge in unemployment. However, some data suggest that it is becoming more difficult for entry-level workers to find software engineering jobs. This is a trend that is in the wave of layoffs in the tech industry.
Google is one of many companies looking to take advantage of the transition to AI-assisted software development. It offers tools ranging from $45 per month to help with code generation and deploy agents that can handle software development tasks. And company It faces widespread competition from rivals such as Microsoft, Openai and Anthropic, as well as AI coding startups such as Replit and Anysphere.
Ryan Salva, who oversees Google's coding tools like Gemini Code Assist, says the “majority” of Google's teams use AI, saying that technology is built into everything from the way the documentation is written in Google's code editor.
“If you're a Google engineer, it's inevitable that you'll use AI as part of your daily work,” he said in an interview with CNN prior to the release of the report.
But just because coders use AI doesn't necessarily mean that everything is finding value in it. 46% of technical experts who participated in the survey said they trusted AI-generated code quality “somewhat”; 23% said they trusted it as “a little”; 20% trusted it as “a lot.” 31% said the AI ”slightly improved” code quality, while 30% said it had “no impact.”
Salva says on a scale of 1-5, one is basic text prediction, and five is the ability to give AI general ambiguous commands. AI software development is between stages 3 and 4. This means that AI can troubleshoot between systems, but still requires human reviews and “many safety nets.”
Adopting AI tools comes at a difficult time for entry-level tech workers. According to the New York Fed, the recent unemployment rate among computer science and computer engineering graduates was higher than in sectors such as ART History and English, with the software engineering role job dropping by 71% between February 2022 and August 2025.
A recent computer science alumni who spoke with CNN earlier this year has generally been optimistic about their future career outlook, but acknowledged that AI is changing the nature of their work. Recent alumni, Julio Rodriguez, said he applied for more than 150 jobs before acquiring the role.
While AI adoption is rapidly advancing, Salva is in a camp for people who believe there is a key part of software development It cannot automate hats, and AI streamlines workers' commonplace parts.
However, he has acknowledged at least some increases in adopting AI. It probably comes from the topic surrounding technology.
“Software development is in the fashion industry. We're all chasing the next style of jeans,” he said. “And when there's a lot of conversation about it, everyone is excited to try something new.”
