More than a third of workers say they’re hoarding expertise because they fear being replaced by the AI agents they’re being asked to train

Details

A Fast Company report reveals more than one in three workers are actively hoarding expertise rather than cooperating with AI agent training initiatives, citing fear of being replaced by the very systems they are helping build. The finding exposes a fundamental conflict of interest at the heart of enterprise AI rollouts. The workers' logic is rational: sharing institutional knowledge, workflows, and judgment with AI agents risks encoding one's own value into a system designed to eliminate that role. One respondent captured the sentiment sharply: 'This is like asking a turkey to vote for Christmas.' The result is deliberate knowledge withholding — slower responses, partial answers, and strategic omissions during training sessions. For enterprise AI teams, this is a structural problem. Agentic AI implementations that depend on internal knowledge transfer face friction precisely among employees with the deepest expertise. Without addressing job security concerns — through retraining guarantees, explicit role preservation commitments, or severance protections — adoption timelines in knowledge-intensive functions may significantly lag projections. Note: Source content is limited to the article headline and subheader; the specific survey or research behind the 'more than a third' figure is not identified in the available text.