When an AI-use contest comes to work, what does it teach people to do?

SHRM’s 2026 survey of more than 5,000 workers says AI competitions are among the approaches workers rate most effective at driving adoption. That is not quite the same as helping people do better work. A contest gives the company an easy scoreboard: prompts, projects, usage. It gives the person with a customer deadline a reason to try the tool where it is least ready, or to keep using it after the answer needs more checking than the task did. If a team runs one, it should reward a useful no as readily as a flashy use: the job the tool was not used for, the mistake caught before it spread, the manual step that disappeared for good. Otherwise it is a scavenger hunt with procurement behind it. What would keep an AI-use contest from becoming a contest in looking busy?