Understanding Global AI Regulation

Countless cases have shown that artificial intelligence (AI) is revolutionising many traditional sectors. With this evolution comes the issue of regulating AI to address its challenges and opportunities. Regulating AI is no simple task; it entails crafting public policy and legislation to promote its beneficial use and mitigate its potential risks.

This process involves managing the regulation of algorithms; several areas of focus include risk management, bias identification, and the explainability of machine learning algorithms in layman’s terms. This is a new challenge, as traditional legal landscapes were not designed to accommodate the rapid development that is characteristic of AI technology. They often find keeping up with AI’s inherent risks and potential advantages daunting.

The Complexities of Global AI Regulation

AI regulation currently garners significant debate worldwide. Entities like the European Union have proactively shaped their approach to this regulation. In 2016, a surge in guidelines centred around AI ethics was observed. This was primarily due to the need to maintain societal control over the rapidly progressing AI technology.

The AI Index, a global project that tracks AI advancements, reveals that the number of laws related explicitly to AI has dramatically increased from one in 2016 to 37 in 2022. High-profile tech industry players, including Elon Musk, the CEO of Tesla and SpaceX, have expressed support for oversight of AI, regardless of the industry’s resulting impacts.

Nevertheless, some professionals believe overly strict regulations could stifle innovation and progress. Instead, they suggest establishing shared norms concerning algorithm testing and maintaining transparency.

Public opinion varies from country to country regarding AI’s potential benefits and possible risks. However, there is an emerging consensus that regulation is crucial to managing these risks. The focus now shifts towards ensuring AI systems are designed strongly emphasising trust and prioritising human needs.

Understanding Hard Law versus Soft Law in AI Regulation 

The landscape of AI regulation includes hard law, such as statutory regulations, and soft law, like industry guidelines. Complex law often faces difficulties adapting to the swift adjustments characteristic of AI technology. On the other hand, while providing the leeway that hard law lacks, soft law may not have the same enforcement power.

The legal minds of Cason Schmit, Megan Doerr, and Jennifer Wagner propose a hybrid approach. They suggest a quasi-government regulator could use intellectual property rights to enforce ethical AI practices. A copyleft licensing could potentially provide a balance between promoting AI innovation and ensuring public safety.

Emergence of Global Guiding Groups on AI Regulation

The need for global oversight in AI development has been echoed since 2017, with multiple parties championing a worldwide governance board to regulate AI progression. In December 2018, France and Canada announced their intention to start an International Panel on Artificial Intelligence (IPAI) backed by the G7 countries. Transitioning into the Global Partnership on Artificial Intelligence (GPAI) in 2020, they aim to ensure AI progress aligns with democratic values and human rights.

Directions for AI Regulation in Selected Nations

While AI regulatory efforts have appeared worldwide, this section will focus on regulations in a few countries: Australia, Brazil, Canada, China, and the EU member states. In Australia, multiple industry groups demanded a unified approach to AI strategy via an open letter in October 2023. Meanwhile, Brazil set grounds for regulating AI by approving the Brazilian Legal Framework for Artificial Intelligence in September 2021.

Canada has been active in AI through its Pan-Canadian Artificial Intelligence Strategy launched in 2017, which aims to foster a new generation of AI researchers and graduates. Its Artificial Intelligence Development Plan guides the state’s AI-related decisions in China. Europe’s strategy is more fragmented, with most countries having national strategies that converge towards a continental strategy.

Evolving AI Regulation 

The lightning-fast developments and potential risks associated with AI demand regulation that evolves as fast as technology. Although approaches vary greatly worldwide, it’s clear that AI regulation is a multifaceted issue that needs much understanding and cooperation. To ensure responsible and ethical use of this powerful technology, laws and policies must focus on the potential risks, biases, and the need to explain AI in a way everyone can understand.

How We Can Help