USA: White House issues Blueprint for an AI Bill of Rights.

The White House Office of Science and Technology Policy issued, on 4 October 2022, a Blueprint for an AI Bill of Rights.

In particular, the Blueprint consists of five principles to guide the design, use, and deployment of automated systems to protect the American public in the age of artificial intelligence ('AI'), each of which are accompanied by a handbook detailing how to incorporate such principles and which include:

• Safe and effective systems: automated systems should be developed with consultation from diverse communities, stakeholders, and domain experts and should undergo pre-deployment testing, risk identification, and mitigation.
• Algorithmic discrimination protections: designers, developers, and deployers of automated systems should take proactive and continuous measures to protect individuals and communities from algorithmic discrimination and use and design systems in an equitable way, including proactive equity assessments and use of representative data.
• Data privacy: privacy should be provided through design choices that ensure protections are included by default, making sure that data collection conforms to reasonable expectations and is only what is strictly necessary for the specific context. Consent should only be used as a legal basis when it can be appropriately and meaningfully given, and requests for consent should be brief and understandable. Sensitive data should enjoy special protections, and unchecked surveillance should be avoided.
• Notice and explanation: designers, developers, and deployers of automated systems should provide documentation clearly describing the role of automation in the overall system, as well as notice that the systems are in the use and an explanation of outcomes, among other things.
• Human alternatives, consideration, and fallback: individuals should be able to opt out from automated systems in favour of a human alternative, where appropriate.

Moreover, the Blueprint discusses how it can be applied, noting the difficulty of doing so due to the speed at which AI develops and the possibility that potential harms of its use occur even with less sophisticated technology. In this regard, the Blueprint poses a two part-test to determine what systems should be within its scope, namely that it should apply to (1) automated systems that (2) have the potential to meaningfully impact the American public's rights, opportunities, or access to critical resources or services.

Lastly, the Blueprint's appendix provides some examples of when automated systems may violate this test, for example automated content moderation tools and surveillance, among others.