Policymakers are stepping up their oversight of AI, often with very different approaches
The US seeks to minimise risks to civil rights and democratic values: On 4 October 2022, the White House released the Blueprint for an AI Bill of Rights, a set of five principles and associated practices to help guide the design, use and deployment of automated systems, while safeguarding the rights of American citizens. This AI framework was informed by and developed in conjunction with various stakeholders, including researchers, policymakers and members of the public, and is intended to support the use of technology and data in ways that reinforce US values and protect people from threats to democracy, equal opportunity and privacy. While many of the concerns addressed in the framework derive from the use of AI, it nevertheless applies to all automated systems that have the potential to meaningfully impact the American public’s rights, opportunities or access to critical resources or services.
The AI framework is underpinned by five principles: The following principles have been developed to respond to challenges posed by the combination and adoption of technology, data and AI throughout everyday life:
Automated systems should be safe and effective, and developed with input from diverse communities and experts to identify potential risks and concerns, including unintended yet foreseeable impacts;
Individuals should not face discrimination by algorithms (i.e. unjustified different treatment based on factors such as race, religion and age), with systems designed and used in an equitable way;
Privacy protections should be embedded by default, with enhanced restrictions in place for sensitive data and heightened oversight for surveillance technologies;
Designers, developers and deployers of automated systems should provide generally accessible plain language documentation, including clear descriptions of the role of automation and implications for individuals; and
The public should be able to opt out from automated systems in favour of a human alternative, where appropriate.
This Bill of Rights will not be enshrined in law: According to the White House, the five principles and associated practices of the Blueprint together form an overlapping set of backstops against potential harms stemming from AI. As the Blueprint is not legally-binding, the US approach currently looks set to align with the actions of countries in Asia Pacific (Australia, Japan, Singapore), which have preferred voluntary rules and existing legislation to address the challenges posed by AI. Elsewhere, policymakers are ramping up efforts to regulate it. The EU is seeking to set the global standard in regulation via the AI Act, and looks set to define and implement rules for AI systems according to a ‘pyramid of risk’. The UK’s approach is also risk-based, but has been pitched as more adaptable and pro-innovation, with decentralisation affording regulators the power to oversee the use of AI within their own sectors.
