Free report·
Faster code, heavier checks
Where AI pressure lands in developer teams
Three-quarters of professional developers (76%) say AI tools are leading managers to expect more output from the same team size.
The 32nd edition of SlashData's Developer survey examines how AI is reshaping software delivery inside engineering teams. It looks at changing delivery expectations, the burden of review and validation, and developers' confidence in the safeguards around AI-assisted work. The central finding is that AI makes code cheaper to write but not software cheaper to deliver: effort saved on writing reappears elsewhere, above all in checking what AI produces. Managers expect more from the same teams, review is becoming a bottleneck, and the people closest to the code are the least sure there is time to get it right. Drawing on a survey of around 11,000 professional developers across 122 countries, complemented by interviews with engineering and technology leaders conducted over the past year, the report shows where that pressure lands and sets out practical steps development teams can take in response.
Key findings
- 76% of professional developers say Al tools are leading managers to expect more output from the same team size, and 68% say developers are under pressure to adopt tools before their value clear.
- 81% say Al reduces repetitive work but increases the need for review and validation, with two in five strongly agreeing.
- 60% of software and architecture professionals say their team has enough time to evaluate AI-assisted work, against 67% of engineering leaders.
Questions this report answers
Does AI reduce the cost of delivering software?
AI makes code cheaper to write but not software cheaper to deliver; the effort saved on writing code reappears elsewhere, above all in checking what AI produces.
Are managers perceived to expect more output from AI-assisted teams?
Yes. 76% of professional developers say AI tools are leading managers to expect more output from the same team size.
Does AI reduce repetitive work?
81% say AI reduces repetitive work but increases the need for review and validation, with two in five strongly agreeing, the strongest conviction of any statement in the survey.
Do developers trust AI-generated code?
Not really. 96% of developers did not fully trust AI-generated code to be functionally correct, yet only 48% always checked it before committing.
Who is least confident there is time to check AI-assisted work?
Software and architecture professionals, the group closest to the generated code, at 60% confidence, compared with 67% of engineering leaders.
Who it's for
- Engineering leaders setting delivery targets for AI-assisted teams
- Software and architecture professionals managing review and validation load
- Technology leaders assessing AI's impact on their operating model