June 27, 2023 Written evidence submitted to Office for Artificial Intelligence on AI regulation Author(s): Minderoo Centre for Technology and Democracy Credit: Scott Webb for Pexels Minderoo Centre for Technology and Democracy has submitted evidence on the proposed pro-innovation framework set out in the UK Government’s AI White Paper. The Office for Artificial Intelligence has requested evidence on the proposed pro-innovation framework set out in the UK Government’s AI White Paper. We have submitted evidence, prepared by Dr Ann Kristin Glenster, Senior Policy Advisor on Technology Governance and Law, Minderoo Centre for Technology and Democracy, University of Cambridge. This submission argues that the framework should articulate the proposed principles and make these mandatory through statutory enactment. Further the submission proposes the adoption of clear transparency and explainability standards, mandatory risk assessments, and a clear guidance as to how human-centric values should be the framework’s governing norm. The UK should adopt a safety-by-design to AI innovation and ensure that regulation is informed not just by industry, but also by civil society and academia in order to foster an environment for responsible AI innovation for the public good. Main points: Legislation should be adopted making the proposed principles clear and mandatory Organisations developing and using AI should have to comport with transparency and explainability standards Organisations developing and using AI should be compelled to publish risk assessments The proposed principles must be human-centric Rather than a ‘duty of having regard’, the UK should adopt a mandatory safety-by-design based approach to the governance of AI innovation Central functions need clear mandates and resources Monitoring must include evidence from civil society and academia The proposed principles should be made specific and mandatory in law, adopting a safety-by-design approach