President Biden is set to issue an executive order outlining the federal government’s initial regulations on artificial intelligence (A.I.) systems. These regulations will include mandatory testing of advanced A.I. products to ensure they cannot be weaponized for biological or nuclear purposes. The order also suggests watermarking photos, videos, and audio created by A.I. to prevent the spread of convincing disinformation, particularly concerning deep fakes.
The U.S. has recently restricted high-performing chip exports to China, aiming to slow down its large language model production. The new regulations will also require cloud service companies to disclose information about their foreign customers.
While these rules are a crucial step in A.I. safety, they primarily focus on future systems and don’t directly address existing A.I. applications like ChatGPT. The order encourages similar regulations among U.S. allies and seeks to establish norms in the technology sector.
The directive instructs various agencies, including the Department of Health and Human Services and the Department of Labor, to create safety standards, streamline A.I. tool purchases, study A.I.’s impact on the job market, and prevent discrimination caused by A.I. algorithms. However, the White House’s authority is limited, and some directives, although significant, might not be enforceable without additional privacy legislation.
The executive order also emphasizes supporting innovation and competition by encouraging the Federal Trade Commission (F.T.C.) to play a more active role in consumer protection and antitrust violations. However, the White House cannot directly regulate the independent agency.
President Biden’s approach strikes a balance between supporting A.I. opportunities in fields like medical and climate research and implementing necessary safeguards against potential abuses. The order also streamlines the visa process for skilled immigrants specializing in A.I., enhancing the U.S.’s position in the global A.I. race.
Crucially, the central regulations related to national security will be outlined in a separate document, the National Security Memorandum, to be released next summer. These regulations, particularly those preventing the exploitation of A.I. systems by foreign nations or nonstate actors, are expected to include classified information.
While there’s eagerness to regulate A.I. technologies, caution has been advised to avoid hasty decisions due to the rapidly evolving nature of these technologies.
For more detailed information, you can access the original article on the source website.
COMMENTS