Key Takeaways for Responsible AI

• Accountability cannot be outsourced to AI: The paper's central finding emphasizes that Product Managers recognize accountability must remain with humans. This is critical for responsible AI adoption—organizations cannot use "the AI decided" as an excuse to avoid responsibility for business outcomes or decisions that affect users.

• PMs are developing frameworks to decide what work is appropriate to delegate: Rather than delegating everything possible, the surveyed PMs (885 total) are thoughtfully assessing which tasks suit AI assistance. This suggests that responsible AI adoption isn't about maximum automation, but about matching AI capabilities to specific work contexts where human oversight remains meaningful.

• Perceived barriers matter as much as benefits: The study identified both perceived benefits and barriers to GenAI adoption among PMs. Understanding what practitioners see as risks—not just opportunities—is essential for building AI systems that teams will use responsibly rather than either over-relying on or rejecting outright.

• Role evolution requires intentional adaptation practices: Rather than AI simply replacing PM work, the research found PMs are developing "adaptation practices" to integrate GenAI into their roles. This suggests responsible AI adoption requires active human decision-making about how roles should change, rather than passive displacement.

• Software development roles are fundamentally changing: The study documents how GenAI is reshaping PM work specifically, highlighting that responsible AI governance must consider impacts across different professional roles, not just technical developers, to ensure accountability structures work across entire teams.