Updated
Updated · The New York Times · Jul 24
Research Institute Identifies 8 Employee Types That Undermine AI at Work
Updated
Updated · The New York Times · Jul 24

Research Institute Identifies 8 Employee Types That Undermine AI at Work

3 articles · Updated · The New York Times · Jul 24

Summary

  • Eight employee types make AI agents harder to use effectively at work, according to a research institute studying how human habits shape AI-assisted performance.
  • Interviews with regular AI users and the agents themselves found the biggest friction often comes from people, not the tools, as workers give outdated, incomplete or poorly framed inputs.
  • Three patterns stood out: the “vague requester” who gives unclear assignments, the “context hoarder” who withholds key information, and the “overdelegator” who leaves decisions to AI and then rejects them.
  • The article argues AI will not compensate for messy workplace behavior the way human colleagues often do, making better management and collaboration skills central to wider AI adoption.

Insights

Is workplace AI really failing because models are weak—or because employees give vague instructions, hide context, and delegate too much?
If AI mirrors your team's habits, what does a bad summary reveal about your management, communication, and trust in automation?