HR leaders in the discussion emphasized that AI should be treated as an enabler, not a master. The key is to start with business and people goals, then decide where AI can help, rather than “chasing the tools.”

Practically, this means:

• Start with outcomes: Define what you want to improve (e.g., faster hiring, better employee experience, more time for managers to coach) before picking tools.
• Focus on roles, not just tools: Several executives argued the real challenge is organizational and role design, not technology. Ask how jobs should be reshaped when AI takes on repeatable tasks.
• Use AI as a partner: Treat AI agents as digital colleagues that handle routine work so people can focus on judgment, creativity, and relationships.

When AI is anchored in clear goals and thoughtful role redesign, it works for your people and your business, instead of people feeling like they’re serving the technology.

Several HR executives stressed that the real opportunity is to reimagine roles from the ground up, not just bolt AI onto current processes.

Examples from the discussion:

• Future-back thinking: One Chief People Officer asked every functional leader to imagine, “If I could blow up my entire HR team and rebuild it from scratch with AI in mind, what would it look like?” The hardest part wasn’t the tech—it was stretching leaders’ imagination.
• Expanding how we see talent: At one company, a monthly contest invites any employee to submit AI agent or use-case ideas. Sales teams, not typically seen as builders, are “running away with it,” revealing people who effectively wear “five functional hats” instead of one.
• HR as a lighthouse: Some leaders believe HR should be a “lighthouse” for AI deployment—piloting AI agents in people processes and then helping the rest of the business adopt them.

The takeaway: redesigning work for AI requires structured imagination exercises, experimentation across functions, and HR taking an active role in modeling what AI-enabled roles can look like.

The leaders in the discussion highlighted a tension between bold AI ambitions and realistic outcomes.

On expectations:

• Some argued for aiming at dramatic productivity improvements (e.g., thinking in terms of “a thousand clinical trials, not one”) to push teams to reimagine what’s possible.
• Others pushed back, saying that incremental, relentless progress is both acceptable and more realistic. For example, consistently improving processes a bit every day can be more credible than promising “thousand-times” returns.

On vendors and partners:

• HR and business leaders are under pressure from shareholders and executives to find the “big bang” that will unlock performance.
• One VP of Talent advised putting real pressure on product partners to justify their roadmaps and clearly explain why they deserve your budget.
• That includes articulating your needs clearly so vendors can align their solutions with your specific value drivers, not just sell generic AI features.

In practice, this means setting balanced expectations—aiming high enough to rethink how work gets done, while recognizing that steady, measurable gains are often what actually show up in the bottom line—and holding vendors to clear, evidence-based value cases.