Encourage, don’t dictate. Rather than outlining precisely how employees should use AI, managers can encourage their direct reports and colleagues to experiment with AI in creative—and ideally transformative—ways. According to Sternfels, one company’s customer service teams were using AI chatbots to draft responses to customers, often dramatically reducing their response times. Some managers, worried about security or governance, shut down those experiments; other managers, meanwhile, studied what made them successful, refined the approaches, and helped scale them. In short: Nurture what’s already growing.

Provide incentives for upskilling. Carrots, not sticks, can increase upskilling. Incentives for AI adoption can include publicly acknowledging employees for demonstrating new competencies and helping others navigate the learning curve. When managers encourage others to learn AI—and along the way, acknowledge that they themselves are still learning—they can lower the psychological barriers to adoption. Companies can also organize innovation competitions, where employees collaborate and submit ideas for how to use AI. The teams that advance to later rounds might receive increased resources, expert support, and leadership exposure—all of which can help boost career development and advancement potential (the top reason Gen Zers cite for taking a new job, by the way).

Foster rapid learning. The most effective managers don’t just experiment more—they experiment better. That means starting with clear, testable hypotheses about how AI can improve ways of working, designing pilots to surface insights (not just “wins”), and embracing quick, small-scale trials that span teams. The goal is fast, cross-functional learning. When managers document why something worked—or didn’t—they help their organizations move faster, make smarter bets, and scale what works.