UK and AI: The United Kingdom’s Approach to Artificial Intelligence Regulation

On the heels of the United Kingdom (UK) announcing its first global summit on artificial intelligence (AI) safety, British Prime Minister Rishi Sunak predicted that London could be the future epicenter of AI regulation. The announcements come as the UK and US have declared a stronger collaboration to tackle AI-related challenges.

The priority of the summit, which is expected to take place in the fall of 2023, is to discuss an internationally coordinated plan to regulate the technology’s growth. Meanwhile, the prime minister and U.S. President Joe Biden continue to meet regularly to coordinate approaches to AI guidelines and regulations.

Unlike the EU, which has adopted a risk-based approach to AI and already approved draft legislation to closely oversee AI, the UK does not plan to enact new laws aimed at governing the technology. Nor does it plan to create new regulatory positions. Rather, the government intends to apply its existing privacy laws and current regulators to oversee AI. Exactly how remains unclear though.

The UK has created a taskforce to collaborate with generative AI models, such as Open AI’s ChatGPT and Google’s Bard, to establish safety and security standards. It will also analyze the strengths of AI to leverage the technology in the country.

The taskforce follows the release of a government policy document in March 2023 which touts the UK’s pro-growth and innovation approach to AI regulation. While the document lacks detail on AI infrastructure for the country, it describes five key principles which regulators will need to abide by when constructing the UK’s regulatory framework: safety, security, and robustness – AI models need to be safe and secure; transparency and explainability – it’s imperative that AI companies be transparent about how their systems work and how they make decisions; fairness – AI systems should not be discriminatory toward individuals or create unfair commercial outcomes; accountability and governance – effective oversight of AI models is necessary; contestability and redress – parties should be able to dispute harmful decisions generated by AI.

Over the next year, the UK is expected to issue guidance to AI companies on how to implement the five principles into their models. It may also pass legislation, granting regulators the authority to ensure the principles are followed.

For additional information, please contact:

Kristopher J. Chandler at kchandler@beneschlaw.com or 614.223.9377.