On July 18, 2022, the UK’s Secretary of State for Digital, Culture, Media and Sport released an overview of its plans to establish a pro-innovation approach to regulating Artificial Intelligence (AI.) The plan envisions a regulatory framework with light-touch regulations that provide the clarity and confidence that’s needed to drive innovation. The Secretary hopes to establish the UK as the best place in the world to found and grow an AI business, while protecting the safety of UK citizens and recognizing that AI is a global ecosystem.
The light-touch means that regulation is targeted to concerns about the actual impact AI has on individuals, groups and businesses. Regulations will be targeted to high risk concerns rather than hypotheticals to avoid placing roadblocks that thwart innovation. Another important facet of the UK’s strategy is to support innovation by providing clear and coherent guidance that's coordinated with the global AI ecosystem. The goal is to use the regulatory leeway that’s been gained from leaving the EU to create a more nimble regulatory framework as a way to attract AI businesses to the UK.
The UK is a leader in the AI ecosystem, ranking first in Europe and third in the word for the amount of investment in AI ($4.65 billion) and number of AI startup companies (49.). In addition to the direct economic benefits of AI investment, there have been tangential gains in other areas that benefit both the economy and society as a whole such as:
- Tracking asbestos cancer tumors;
- Helping people facing fuel shortages and
- Improving animal welfare on dairy farms.
With these gains come concerns over privacy rights and safety that require the government to set standards to protect its citizens. Another issue is how much trust should be put in algorithms to protect critical infrastructure and under what circumstances backups need to be put in place. There are many organizations in the UK and the EU that are developing protocols, but these diverse solutions must be properly coordinated to ensure the growth of the AI industry in the UK is not slowed down by excessive regulation.
As AI technology becomes more sophisticated, it will also become increasingly difficult to regulate. This is because the goal of AI is often to go beyond what is discernible to humans to perform tasks such as diagnosing illnesses and completing missing passages of ancient texts. The regulatory challenge intensifies as the decisions made by AI become increasingly difficult for humans to understand and explain. For example, self-driving vehicles still require a driver to be alert and ready to take over if needed, operating in a manner similar to the autopilot functions on a plane. When more sophisticated self-driving AI is developed, motor vehicle manufacturers may want to advertise that no driver control is necessary and it will be up to regulators to decide whether these claims are justified.
The rapid and sometimes unexpected transformative impact of AI development makes setting forth a predictable government response crucial for the growth of the industry in the UK. The pro-innovation approach seeks to achieve this goal by incorporating these four core principles:
- Context-specific regulations targeted to the risks arising from a particular context rather than general concerns;
- Risk-based approach to regulating high risk applications, avoiding regulations that address low risk and hypothetical situations;
- Clear, simple, stable and predictable regulations that the AI industry can rely on and
- Preference for light-touch voluntary measures or guidance over rigid mandatory legislation.
Over the coming months, the UK’s Secretary of State for Digital, Culture, Media and Sport will be seeking input from consumer groups, business interests and other stakeholders about this approach and how to best refine and implement it. This process will include a thorough analysis of gaps in coverage that may require more targeted solutions, the roles of various government agencies in developing and implementing specific regulations and developing a system to oversee the regulations and monitor new developments. The full text of this report is available here and the results will be published in a White Paper in late 2022.
