With policymakers moving quickly on AI safety, from the EU AI Act to the US AI Action Plan, new research argues that uniform, one-size-fits-all rules could unintentionally entrench the largest AI firms and narrow the space where start-ups innovate.

In this episode, Dr Pavel Kireyev, Assistant Professor of Marketing in the Department of Management at LSE, discusses his recent research, User preferences for large language model refusals: implications for moderation and market structure.

Dr Kireyev argues for asymmetric regulation of AI firms, enforcing stricter guardrails for the largest firms and carefully scoped flexibility for smaller and open‑source providers, thereby offering a better balance between safety, innovation, and competition. 

Dr Kireyev conducts research with data-driven companies and studies innovative marketing technologies, platforms, and marketplaces. He uses modern quantitative methods to uncover how organisations can effectively coordinate their pricing and advertising strategies across multiple platforms, benefit from new resources such as crowd intelligence, and manage data to improve decision-making and market design in multi-sided marketplaces.