Japanese companies are building data foundations and policy frameworks to let AI agents handle HR tasks—but success depends on having both clean employee data and clear governance rules in place first.

What happened: Asahi Kasei is gradually deploying AI agents in parts of the HR business but has not yet rolled them out company-wide, citing gaps in data quality and employee AI literacy. PayPay has mapped HR work across two dimensions—strategic versus operational, and bespoke versus standardized—to decide where to build custom AI tools versus use existing systems. PayPay also rewrote its privacy policy to pre-approve employee data use for AI, allowing faster development without repeated legal checks.

Why it matters: Both companies see AI agents as a way to automate and improve HR decisions like internal job matching and career support, but they found that success requires clean, detailed data (skills, career histories, even calendar logs) and clear policies governing how employee information is used. Asahi Kasei notes that current data alone is not sufficient for high-quality matching; data governance must happen in parallel with system design for AI to work.

What to watch: The central tension both companies are navigating is whether AI can handle 'seeing people'—the judgment-heavy work of understanding individual strengths and fit—or whether that remains uniquely human. The degree to which companies can solve the data and policy puzzle will determine whether AI augments HR or displaces it.