AI’s time savings are disappearing into code review

An Atlassian survey reported by LeadDev finds developers saving ten hours a week with AI fell from 68% to 25%, as reviewing AI-written code takes up the gains.

Michael Hill, managing editor of LeadDev, reports on Atlassian's 2026 State of Developer Experience survey of 1,155 developers and 405 engineering leaders. Although 87% of the developers say AI tools help them write and ship code faster, 72% report the same or more friction than before AI. Most AI deployments still focus on writing code rather than on how work moves across a team.

• In 2025, 68% of developers said AI saved them at least ten hours a week; this year 25% say so, and 65% regularly review, debug and correct AI-generated code.
• The largest sources of friction barely changed: finding information at 51%, meetings at 50%, context switching at 45%, technical debt at 44% and unclear requirements at 40%.
• Only 17% of developers and 23% of engineering leaders say their organisation has a clearly communicated and implemented AI strategy.
• Of developers, 60% say their organisation treats AI usage as a productivity signal, while 42% of leaders believe they have the right measure to separate real gains from activity.

An engineering team can check whether faster code generation has simply moved its queue to review, and measure delivery rather than prompts; this is an inference and is not the author's claim. The figures come from a survey by Atlassian, which sells tools to developers.

1 November 2026: Teams putting AI agents to work decide in advance which steps a person must still take, and that person carries more review and a greater duty to push back.