Atlassian Report Says AI Speeds Coding but Not Software Delivery
A new Atlassian survey finds 87% of developers say AI helps them write and ship code faster, but 72% report unchanged or greater friction across the software lifecycle, with only 17% saying their employer has a clear AI strategy.
The survey also found that longer-term optimism about AI's time savings is falling. Only 25% of developers say they save 10 or more hours a week when using AI, down from 68% in 2025, according to Atlassian. The report concludes that AI is shifting bottlenecks downstream rather than removing time constraints across the delivery process.
Human bottlenecks remain prominent. Half of developers cite difficulty finding information (51%) and meeting frequency (50%) as problems, Atlassian said. Andrew Boyagi, Atlassian's customer CTO, said: "Faster code is not the same as better software delivered sooner. If teams still struggle to find information, clarify requirements and move work through the delivery process, speeding up coding alone won’t translate into business outcomes."
Many developers also question how AI is being adopted. Three in five say that merely using AI is treated as a positive metric, even though its real-world impact is not measured in the same way, according to the report. Organizations with an effective AI strategy were in a better position to reduce friction, with twice as many workers reporting positive effects, but only 17% of developers said their employer has a clearly communicated and implemented AI strategy.
The findings suggest more than four in five organizations still lack a clear AI strategy. Atlassian’s report calls for clearer, top-down strategies as the industry moves beyond early experimentation with AI, arguing that use cases need to be identified and managed with their effects on the whole development process in mind.