Humans in the loop: how software teams are learning to trust AI (www.techradar.com)

Recent insights highlight how software engineering teams are learning to effectively integrate AI into their workflows, despite initial concerns over risks associated with AI-generated code. The 2025 DORA survey from Google reveals that AI can enhance existing quality processes, leading to higher quality outputs. Currently, 76% of enterprise teams are scrutinizing AI-generated work with the same rigor as human-produced code, demonstrating a commitment to maintain engineering standards. This cautious but progressive approach encourages teams to treat AI as a virtual teammate, retaining human oversight in critical decision-making stages, especially during the potentially high-risk release phase. The gradual adoption of AI seems to promote a disciplined framework within software development, where engineers can redirect their focus to essential aspects like code review, testing, and risk assessment. As organizations strive for a balance between speed and diligence, those that successfully integrate AI while adhering to robust engineering practices are likely to see significant benefits. The shift exemplifies a nuanced understanding of AI's role in development—placing human judgment at the forefront while leveraging AI to alleviate mechanical tasks. This iterative process fosters trust in AI systems, indicating a promising future for AI-supported software delivery.