Currently, the U.S. does not have a federal framework to regulate AI. The Algorithmic Accountability Act was introduced in the U.S. Senate and House of Representatives on February 3, but even if it passes, it won’t be able to take effect for a few years. In the meantime, states such as Alabama, Colorado, Illinois and Mississippi have passed laws regulating AI, and others are considering bills. The Electronic Privacy Information Center (EPIC) recently published “The State of State AI Policy,” a summary of AI-related legislation that states and cities considered between January 2021 and August 2022.

To better understand the emerging opportunities, challenges, and implications resulting from AI developments, the GAO (U.S. Government Accountability Office) in 2017 hosted an AI forum in Washington, DC. While this raised the issue of the adequacy of the laws and regulations then in place, no federal framework is in place today, the Algorithmic Accountability Act, which by the way is aligned with the GDPR, could provide for that.

For now, States and cities have adopted different AI regulatory policies that “The State of State AI Policy” highlighted.

Laws at the state and city level

Seven laws have been enacted to regulate various AI applications and activities.

Facial recognition

Two states and two cities have restricted facial recognition.

• In Alabama, law enforcement agencies (LEAs) are no longer allowed to use facial recognition to determine probable cause during a criminal investigation or attempted arrest. When LEAs are seeking to establish probable cause, the law only allows LEAs to use the match results of facial recognition technology in conjunction with other lawfully obtained information and evidence.
• Colorado has instituted a similar law that prohibits state and local government agencies from using it to identify, monitor, or track individuals without a warrant and also prohibits the technology in all public schools. On the other hand, the law requires state and local agencies that use or intend to use a facial recognition service to subject decisions that have legal effect to meaningful human review and to conduct periodic training for those who use it.
• Baltimore City, Maryland, has banned all private and non-police officials from using facial recognition within the city limits.
• Bellingham, Washington, bans law enforcement from using facial recognition or predictive policing tools. Bellingham residents voted to prohibit the city from acquiring or using facial recognition technology or contracting with a third party to use facial recognition technology on behalf of the city. The measure also restricts the use of illegally obtained data in police or court proceedings.

Automated decision making

Two states and one city have restricted automated recruitment.

• Vermont has created an agency to review state uses of AI.
• Illinois employers that rely solely on AI analysis of video interviews are required to report the race and ethnicity demographics of applicants not selected for personal interviews and those hired.
• New York City employers using these tools are required to inform job applicants and verify these tools before using them.
• Artificial intelligence education: Mississippi passed a law directing the state’s education department to produce an artificial intelligence and machine learning program for public schools.

Impacts of AI

Two states have implemented government oversight of the technology.

• Alabama created a council to advise legislators on the use and development of automation in the state.
• Illinois has formed a task force to forecast the impact of AI and other new and emerging technologies on employment, wages, and skill requirements for jobs in the state.

Artificial Intelligence Education

Mississippi passed the “Mississippi Computer Science and Cyber Education Equality Act” which directs the state department of education to implement a K-12 computer science curriculum that includes AI and machine learning courses.

Legislation in the works

Thirteen other bills are currently underway in Washington DC, California, Georgia, Maryland, Massachusetts, New Jersey, New York, and Rhode Island to study the impacts of AI. California lawmakers, meanwhile, are proposing mandatory processes to minimize algorithmic bias, their Hawaiian brethren a tax credit for AI companies.

Translated from Où en est la réglementation de l’intelligence artificielle aux USA ?