To address concerns about potential abuse of artificial intelligence (AI) systems, the White House has announced a set of principles called the Blueprint for an AI Bill of Rights. These five principles will guide the design, use, and deployment of automated systems in both the public and private sectors, particularly those that have the potential to significantly impact the American public’s rights, opportunities, or access to critical resources or services.
The five principles are as follows:
- Protection from unsafe or ineffective systems
- No discrimination by algorithms, and equitable use and design of systems
- Protection from abusive data practices with built-in protections, and agency over how data is used
- Awareness of the use of automated systems, and understanding of how they contribute to outcomes
- Option to opt-out, and access to quick problem resolution
While automated systems have brought numerous benefits, such as improved efficiency in farming and disease identification, their use in hiring and credit decisions has been found to reproduce unwanted inequities or embed new bias and discrimination. In addition, social media data collection can threaten privacy and opportunities, often without individuals’ knowledge or consent.
The Blueprint for an AI Bill of Rights is a guide for society to protect individuals from these threats and ensure that technologies reinforce our highest values. It is not legislation or regulation but a guide for the public and private sector.
The National Science Foundation will also add $140 million in funding to launch seven new National AI Research Institutes.
Furthermore, the U.K. government has asked regulators to support the innovative development and deployment of AI against five overarching principles, including safety, transparency, and fairness. The U.K. Competition and Markets Authority will review how AI foundation models are developing, assessing the conditions and principles that will best guide their future use.
Canada is also currently dealing with the proposed Artificial Intelligence and Data Act (AIDA) to govern the use of AI in high-impact systems. However, critics complain that the act lacks necessary details that will only be filled in by regulations after it passes, and they also take issue with the fact that the proposed AI data commissioner will report to the Minister of Innovation and not be independent.