Task Expansion Triggered by AI
Researchers from UC‑Berkeley and Yale observed that, after eight months of embedded AI use in a 200‑person tech firm, traditional role boundaries dissolved. Product managers began writing code, and researchers took on engineering tasks, making previously siloed work feel feasible.
Ripple Effect on Engineers
Engineers found themselves reviewing, correcting, and coaching colleagues—a phenomenon participants called “vibe‑coding.” The automation of one person’s task often generated additional work for another.
Surge in Multitasking
AI gave the impression that tasks could run in the background, prompting employees to juggle multiple workstreams simultaneously. The promised productivity gains turned into constant attention‑switching and longer task lists.
Self‑Reinforcing Cycle Leading to Burnout
Researchers defined a loop: AI makes tasks easier → workers do more of those tasks → reliance on AI grows → the cycle repeats, ultimately causing burnout.
- “Several participants noted that although they felt more productive, they did not feel less busy, and in some cases felt busier than before.”
- Workers are being laid off, and those who remain face stretched capacities.
Survey Evidence of Widespread Fatigue
A DHR Global survey of 1,500 corporate professionals found 83% experiencing burnout, with overwhelming workloads and excessive hours as top drivers. In 2024, Upwork Research Institute reported 77% of AI‑using employees said AI decreased productivity and increased workload.
The most in‑demand skills this year are AI‑related, indicating a market shift toward managing AI tools rather than reducing work.
Senior‑Level Burnout Gap
Burnout rates differed by seniority: 62% of associates and 61% of entry‑level workers reported burnout, compared with 38% among C‑suite leaders.
Recommendations for Sustainable AI Integration
- Implement structured pauses before major decisions.
- Sequence work to limit context‑switching.
- Protect dedicated time for human connection.
- Monitor workload metrics to detect early signs of cognitive fatigue.
Without these practices, AI‑assisted work tends toward intensification rather than contraction, jeopardizing decision quality and long‑term sustainability.