Guardrails versus Gatekeepers: Understanding Product Managers' Ethical Decision-Making in Generative AI

Uncertainty and diffused responsibility are major barriers: Product managers struggle to take ethical action on generative AI because they're unclear about what "responsible AI" means in practice and unsure whose responsibility it is to ensure ethical use—highlighting a need for clearer organizational guidance and accountability structures.

Leadership commitment dramatically increases ethical action: Organizations with strong leadership commitment to responsible AI practices make ethical actions up to 14 times more likely among product managers, suggesting that top-down support is essential for translating ethical principles into actual product decisions.

Product managers occupy a critical but overlooked role: While engineers and ethicists have received significant attention in AI responsibility discussions, product managers—who function as key decision-makers in product teams—have been largely neglected despite their power to shape how generative AI is deployed.

Two-tiered approach to ethical action exists: Product managers employ both low-resource individual actions (that don't require organizational support) and high-resource collective actions (that do require organizational incentives), suggesting that responsibility for ethical AI cannot fall entirely on individual initiative.

Organizational incentives must align with ethical commitments: The research reveals that ethical policies often become decoupled from actual practices when profit motives and organizational incentives don't support responsible action, indicating that companies must restructure how they reward and measure product management success.