Manyika calls for shared responsibility in AI regulation

Shared responsibility for AI safety

Google senior vice president James Manyika said AI risks are real, regulation is necessary and responsibility must be shared across industry, government and society. He made the case in an interview that aired on October 2, 2026.

Manyika said companies building frontier AI models should be responsible for making those systems safe. He also assigned roles to deployers, users and policymakers, arguing that no single group can own the entire safety problem.

Industry action alongside government

Manyika opposed the idea that technology companies should wait for President Trump to regulate AI. He argued that voluntary industry efforts should proceed in parallel with the rules governments eventually establish.

Google, OpenAI and Nvidia are among the major technology firms that recently signed a voluntary accord. The agreement focuses on robust internal safety processes and independent auditing. It has no legal enforcement mechanisms, meaning no regulator is positioned to fine a company that fails to meet its commitments.

Balancing safety and progress

Throughout 2025 and 2026, Manyika has argued that AI needs regulation but should avoid rules that block useful progress. He supports balanced, enabling frameworks that address genuine harms while allowing AI to deliver benefits in areas such as health and infrastructure. He also co-directs Google DeepMind Institute.

His position presents leading technology companies as proactive on AI safety, while the lack of enforcement leaves critics focused on the accountability gap. The next issue is whether voluntary commitments can develop into enforceable standards and whether companies or governments will define the baseline.

Sources: Bloomberg Via Crypto Briefing