A Blueprint for an AI Bill of Rights was recently published by the White House Office of Science and Technology Policy. Five principles are outlined in the Blueprint for designing, deploying, and using automated technologies to safeguard the American public in the age of artificial intelligence (AI).
These five principles consist of:
- Safe and effective systems. Automated systems should be built in collaboration with various communities, stakeholders, and domain experts, and they should go through pre-deployment testing, risk identification, and mitigation;
- Algorithmic discrimination protections. Designers, developers, and deployers of automated systems should take proactive and ongoing steps to protect individuals and communities from algorithmic discrimination and use and design systems in a fair way, such as by doing proactive equity assessments and using representative data;
- Data privacy. Privacy should be offered through design decisions that ensure safeguards are incorporated by default, ensuring that data gathering adheres to reasonable expectations and is limited to what is strictly necessary for the specific context. Consent should be utilised as a legal foundation only when it can be supplied effectively and meaningfully, and consent requests should be brief and comprehensible. Sensitive data should be safeguarded, and unregulated surveillance should be avoided;
- Notification and explanation. Designers, developers, and deployers of automated systems should include documentation that clearly describes the function of automation in the entire system, as well as notification that the systems are in use and an explanation of outcomes, among other things; and
- Human alternatives, consideration, and fallback. Where appropriate, individuals should be allowed to opt out of automated systems in favour of a human option.
The AI Bill of Rights Blueprint is a guide for a society that protects all people from these challenges and uses technology to reinforce ideals. Responding to public experiences and informed by researchers, technologists, advocates, journalists and policymakers, this framework is accompanied by a technical companion – a handbook for anyone seeking to incorporate these protections into policy and practice, including detailed steps toward actualising these principles in technological design.
These principles aid when automated systems harm public rights, opportunities, or access to crucial needs. Moreover, the Blueprint examines how to apply AI, stressing its difficulty due to its rapid development and the likelihood of harm even with less advanced technologies.
While many of the problems in this paradigm arise from AI, the technical capabilities and definitions of such systems evolve with innovation, and the potential downsides of their use exist even with less complex instruments.
This framework uses a two-part test to establish scope. This approach applies to (1) automated systems that (2) could harm Americans’ rights, opportunities, or access to key resources or services. These rights, opportunities, and access to key resources and services should be enjoyed equitably and completely protected, despite the increasing role of automated technologies.
Moreover, the framework outlines the precautions that should be applied to all automated systems that have the potential to have a significant impact on the exercise of a) Civil rights, civil liberties, and privacy, including freedom of expression, voting, and protections against discrimination, disproportionate punishment, unlawful monitoring, and abuses of privacy and other freedoms in both the public and private sectors.
In addition to b) Equal opportunities, including access to education, housing, credit, employment, and other services on an equitable basis; or c) Access to essential resources or services, including healthcare, financial services, safety, social assistance, accurate information about goods and services, and government benefits.
