On March 29, 2023, the UK government released a long-awaited white paper detailing how the UK will regulate AI. As expected, the UK reaffirmed its stance from her July 2022 National AI Strategy and opted for an innovative approach to AI regulation. The government has introduced a framework containing five cross-cutting principles to guide and inform the responsible development and use of AI.
1) Safe, secure and robust.
2) Appropriate transparency and explainability.
3) Fairness.
4) Accountability and Governance.
5) Competition and Indemnification.
Given the lack of consensus to define AI in practice, this white paper focuses on the distinct characteristics of AI applications in specific sectors, use cases, and developments. The UK intends to allow regulators to determine the scope of AI within the context of its use. In particular, the White Paper outlines that the regulatory regime must reflect these essential characteristics.
• Promoting Innovation – Enabling rather than stifling responsible innovation.
• Balance – avoid unnecessary or disproportionate burdens on companies and regulators.
• Trustworthy – address real risks, increase public confidence in AI, and drive and facilitate AI adoption.
• Adaptability – Help people adapt quickly and effectively to meet new opportunities and risks as AI technology evolves.
• Clarity – Stakeholders in the AI lifecycle, including companies using AI, can easily know what the rules are, who applies to them, who enforces them, and how to comply with them.and
• Collaborative – Encourage governments, regulators, and industry to work together to foster AI innovation, build trust, and ensure that the voice of the public is heard and considered.
Holding regulators accountable can be effective in ensuring that regulatory principles are properly applied. Sector-specific expertise may give regulators better equipped to assess the scope of specific AI uses. Additionally, regulatory discretion leans toward a light-touch approach that avoids the use of restrictive laws and increases flexibility for innovation, which may please most AI stakeholders.
However, the UK approach has its drawbacks and differs significantly from the EU’s restrictive and risk-based regulatory approach. Especially given the globalized nature of AI, companies that develop and adopt AI applications are likely to require compliance with EU law anyway. The intent of the UK AI approach may therefore be eroded somewhat.
Overall, this white paper aims to reaffirm the UK’s intention to create an environment for AI innovation. Given that this white paper is a public consultation, there is room for the UK to refine its approach to AI regulation to ensure that an innovation-promoting approach is delivered over the long term.
