01 · The problem
AI scales output. Human oversight capacity doesn’t scale with it.
Humans remain accountable for work AI creates. That doesn’t change as AI activity increases. It just gets harder to honour.
Where it breaks
Review becomes the bottleneck.
Approval becomes rubber-stamping.
Accountability stays human regardless.
A human in the loop is not meaningful oversight.
37%
of DevSecOps professionals would trust AI to handle daily work tasks without human review.
Source · GitLab / The Harris Poll · Global DevSecOps Survey · n=3,266
Day to day
What oversight pressure looks like from the inside.
- 01
Two-line requirements from leadership expand into uncosted complexity by the time they reach the team.
- 02
The person closest to overload is often the last to say so.
- 03
A slowdown becomes visible in a retrospective, months after it started.
- 04
Wellbeing tools get rolled out, then quietly deleted, because nobody trusted what they were really measuring.
Unattributed scenario patterns from FreQNC discovery research, 2026. No participant is named or quoted.