While AI tools help developers write and ship code faster, a new report shows they haven't yet solved productivity challenges across the full software development lifecycle, with many teams still facing friction in crucial non-coding activities.
- 87% of developers say AI speeds up coding but 72% see persistent lifecycle friction.
- Only 25% now save over 10 hours weekly with AI, down from 68% in 2025.
- Just 17% report clear employer AI strategies correlating with reduced friction.
What happened
Atlassian's latest survey of software developers reveals a paradox where AI-assisted coding is undeniably speeding up the generation of code, with nearly nine in ten acknowledging faster output. However, this improvement does not extend across the broader software development lifecycle. Most developers reported that other stages, such as gathering information and managing requirements, remain problematic or even more difficult despite faster code production.
The survey also notes a significant decline in time savings attributed to AI, with only a quarter of developers saving more than 10 hours per week currently, compared to nearly 70% in 2025. This indicates that while AI can quicken specific coding tasks, it shifts productivity bottlenecks rather than eliminates them.
Why it matters
The findings underscore the critical insight that faster code creation does not necessarily lead to faster or better overall software delivery. Developers face persistent challenges with information accessibility and excessive meeting demands, which impede their ability to move work efficiently through development pipelines. Without addressing these systemic issues, gains in coding speed alone will fail to produce meaningful business outcomes.
Furthermore, the report highlights a gap in organizational preparedness, as only a minority of companies have implemented clear and communicated AI strategies. Where such strategies exist, developers report fewer workflow frictions, suggesting that strategic adoption is essential to unlocking AI's full productivity potential in software development environments.
What to watch next
Moving forward, companies and technology leaders should focus on creating comprehensive AI adoption plans that go beyond coding assistance. This includes prioritizing solutions that improve information sharing, coordination, and requirement clarity across teams to tackle the root causes of workflow friction and realize faster software delivery.
Additionally, monitoring how AI tools evolve to support other parts of the software lifecycle—such as testing, deployment, and project management—will be crucial. Stakeholders should watch for developments in AI integrations that more holistically address these broader productivity challenges, as well as increased organizational emphasis on communicating and measuring the real-world impact of AI implementations.