AI tools impact developer productivity and open-source maintenance

A recent Atlassian study, ‘State of Developer Experience 2026’, reveals that while 87% of software developers agree that AI tools help write and deploy code faster, these gains are not translating into significant time savings. The percentage of developers reporting at least ten hours saved per week dropped from 68% the previous year to 25%. This decline is attributed to increased friction in the development cycle, with 65% of developers noting additional effort required to review AI-generated code.

Other factors slowing progress include information searching (51%), meetings (50%), frequent context switching (45%), technical debt (44%), and unclear requirements (40%).

Simultaneously, the open-source community is facing challenges from an influx of reports and ‘AI slop’. Daniel Stenberg, developer of the Curl project, reported an ‘avalanche of vulnerability reports’ that has left maintainers exhausted and overwhelmed. To manage the workload, the Curl team will restrict its private HackerOne tracker to only high or critical vulnerabilities starting October 26, moving medium and low-level issues to public GitHub repositories to reduce the burden on central maintainers.