Workers Sabotage AI Mandates Across Industries as 600% Slower Tools Fuel Resistance
Updated
Updated · Aftermath · Jul 30
Workers Sabotage AI Mandates Across Industries as 600% Slower Tools Fuel Resistance
3 articles · Updated · Aftermath · Jul 30
Summary
Employees across gaming, tech, utilities, finance and other sectors are quietly resisting employer AI mandates by faking usage, feeding chatbots junk prompts, skipping recorded meetings or submitting unedited AI output to expose errors.
Managers often track prompt counts rather than work quality, creating easy loopholes: one worker won an "AI Champion" cash prize with useless prompts, while others asked Copilot trivia questions to satisfy performance metrics.
Resistance is also driven by poor results and added risk. Workers described AI translations that had to be redone by humans, an "agentic" tool that took 600% longer, and security, copyright and confidentiality concerns.
Some supervisors are shielding teams by accepting token compliance or citing human-review rules, but several workers said AI adoption is becoming standard in Western AAA studios and is increasingly tied to evaluations and job security.
The accounts suggest a widening gap between executive pressure to show AI adoption and frontline skepticism that the tools improve speed, accuracy or business outcomes.