On February 10, 2020, the UK Government’s Committee on Standards in Public Life* (the “Committee”) published its Report on Artificial Intelligence and Public Standards (the “Report”). The Report examines potential opportunities and hurdles in the deployment of AI in the public sector, including how such deployment may implicate the “Seven Principles of Public Life” applicable to holders of public office, also known as the “Nolan Principles” (available here). It also sets out practical recommendations for use of AI in public services, which will be of interest to companies supplying AI technologies to the public sector (including the UK National Health Service (“NHS”)), or offering public services directly to UK citizens on behalf of the UK Government. The Report elaborates on the UK Government’s June 2019 Guide to using AI in the public sector (see our previous blog here).
Public-sector organisations are increasingly exploring ways of using AI to deliver better public services. The Committee found that, in the UK, healthcare and policing currently have the most developed AI programmes, having benefitted from sector-specific codes (such as the Code of Conduct for Data-Driven Health and Care Technology in the healthcare context, available here). The Report contends that implementing clear ethical standards around AI will accelerate adoption of innovative technologies by building user trust in them.
The Report sets out recommendations to the UK Government in its roles as policymaker and regulator, alongside recommendations to providers, both public and private, of public services. Although many of these recommendations will be primarily relevant to public officials, several could impact companies that either supply AI products or services to public-sector customers in the UK, or that provide public-sector services using AI on behalf of UK public-sector organisations.
Recommendations to the UK Government
Of the Report’s recommendations to UK policymakers and regulators, the following are likely to be of particular interest to companies providing AI technologies to UK public-sector customers:
- Procurement Rules and Processes. The Government should leverage its market purchasing power to ensure that private companies develop public-sector AI solutions that meet appropriate public standards.
- Crown Commercial Service’s Digital Marketplace. The Crown Commercial Service should introduce practical tools to assist public bodies and those delivering public-facing services to easily identify AI products and services that meet their ethical requirements.
- Impact Assessment. The Government should consider how an AI impact assessment requirement could be integrated into existing processes to evaluate the potential effects of AI on public standards; such assessments should be mandatory and should be published. If adopted, this recommendation could lead UK public-sector customers to require documentation from suppliers that enables them to conduct such assessments.
- Transparency and Disclosure. The Government should establish guidelines for public bodies about the declaration and disclosure of AI systems. To the extent this includes disclosures about the design or operation of the system, this too could lead public-sector customers to impose additional documentation demands on suppliers of AI technologies.
Recommendations to providers, both public and private, of public services.
The Report also offers seven recommendations to front-line providers, both public and private, of public services:
- Evaluating Risks to Public Standards. The potential impact of a proposed AI system on public standards should be assessed at a project design stage, ensuring that the design of the system mitigates any identified risks to these standards. This review should be renewed upon any substantial change to the design of an AI system.
- Diversity. Issues of bias and discrimination should be tackled by ensuring account is taken of a diverse range of behaviours, backgrounds and points of view.
- Upholding Responsibility. Responsibility for AI systems should be clearly allocated and documented, so that operators for AI systems are able to exercise their responsibility in a meaningful way.
- Monitoring and Evaluation. Providers of public services should continually monitor and evaluate AI systems to ensure they always operate as intended.
- Establishing Oversight. Oversight mechanisms should be established that allow for AI systems to be properly scrutinized.
- Appeal and Redress. Citizens should always be informed of their right and method to appeal against automated an AI-assisted decisions.
- Training and Education. Employees working with AI systems should undergo continuous training and education.
The Report concludes that the UK’s regulatory and governance framework for AI in the public sector remains a work in progress. Although the Committee commends the work being done by the Office for AI, the Alan Turing Institute, the Centre for Data Ethics and Innovation (CDEI), and the Information Commissioner’s Office, it notes that there is an urgent need for guidance and regulation on the issues of transparency and data bias in particular. The Committee considers that the UK does not require a specific AI regulator, but that existing relevant regulators should be assisted by the CDEI, acting as a “regulatory assurance” body. The Committee emphasises that public-sector bodies must comply with existing UK law on data usage and should implement clear, risk-based governance frameworks for their use of AI. We can expect that public bodies, such as the NHS, will seek to flow down these requirements to companies providing AI technologies.
The use of AI in the public sector is an issue of increasing debate and focus, particularly in Europe. The team at Covington will continue to monitor developments in this area.
*The Committee on Standards in Public Life is an advisory non-departmental public body that advises the UK Prime Minister on ethical standards across the whole of public life in England. It monitors and reports on issues relating to the standards of conduct of all public-office holders, and its recommendations inform policy decisions of the UK Government.