Article by Daniel Lloyd, Partner, Technology and IP Team, TLT LLP
 

A Fragmented yet evolving landscape
 

The global AI regulation landscape is fragmented and rapidly evolving. After the initial optimism for enhanced cooperation among global policymakers, the reality is that different regions are adopting varied models, ranging from policy statements to legislation. Despite this fragmentation, a common direction is emerging, focusing on minimising AI risks and establishing structures for safe and ethical AI development. But the distinct regional differences now emerging are worth considering. Let’s look at each region in turn. 

The EU: In regulatory overdrive with the EU AI Act: A Brief Overview
 

The EU AI Act, officially published on July 12, 2024, represents a landmark regulation aimed at fostering responsible AI development and deployment across the European Union. This Act introduces a harmonized legal framework designed to ensure AI systems are trustworthy, human-centric, and aligned with fundamental rights and values. At the heart of the Act is a risk based approach that categorises AI systems based on their risk levels, from minimal to unacceptable risk, with corresponding regulatory requirements. The Act also introduces requirements in respect of AI deployments for: Transparency and Accountability; Ethical Standards and Safety and Environmental Protection. The Act is also extra territorial in its reach and in theory applies to any entity deploying AI applications within the EU. 

The extent to which the Act may discourage big American Tech from deploying AI within the EU, or stifle innovation within the EU remains to be seen. But the Act at least creates certainty for business, consumers and developers alike, based on trust and public safety being the most important policy considerations. 

The USA: Fragmentation and confusion on AI from the White House
 

The exact opposite appears to be the case in the USA. Prior to Trump the USA was actively developing a regulatory framework for AI, although at a federal level it lacked and still lacks a comprehensive federal AI Act. Of the federal initiatives to survive from the Biden era it still appears to be the case that Federal Agencies like the Federal Trade Commission (FTC) and the National Institute of Standards and Technology (NIST) continue to develop AI standards and guidelines. At state level a more fragmented picture emerges with many different states adopting their own laws. States like California, Colorado, and Illinois are leading in AI regulation. For instance, the Colorado AI Act of 2024 uses a risk-based approach similar to the EU AI Act, focusing on transparency and risk mitigation. We don’t know yet what if anything the USA will do at a federal level to create a regulatory framework for AI but we should assume that it will be more focused on encouraging innovation compared to the EU AI Act. 

And what of China? 
 

China is working on a comprehensive AI law that aims to cover a wide range of AI-related activities. This draft law, which is still under development, focuses on promoting technological innovation, regulating AI product and service development, and safeguarding national security and public interests. China's approach to AI regulation emphasizes the importance of ethical considerations, transparency, and risk management, aligning with global trends while also addressing specific national priorities in defence of communist party rule. Only in China are there laws in place stating that LLMs like Chat GPT cannot ingest material which is contrary to fundamental socialist values. And the leading Chinese generative AI App, Deep Seek, is well known among other things for not being able to give an account of recent events in Chinese history that conflicts with the official version. By way of illustration if you ever get the chance - just ask Deep Seek about what happened in Tiananmen Square in June 1989. 

What should the UK approach be?
 

The UK is the third largest AI economy in the world, after China and the USA. It will be challenging for the current administration to strike the right balance between aligning with the EU while maintaining an open relationship with the USA. A long period of consultation should be anticipated as the government will need to pay a lot of attention to striking the right balance between innovation and fundamental human rights. 

Find out more: Visit TLT LLP

This article was first published in the Spring edition of Society Matters magazine