President Biden issues an executive order intended to curb AI risks

Privacy & Compliance

US President Joe Biden is looking to use a new executive order that addresses the topic of risks posed to consumers, workers, minority groups, and national security by artificial intelligence. The order will require AI developers behind systems that pose risks to the US economy, public health, national security, or safety to share the results of safety tests with the government, as per the Defense Production Act, before releasing them to the public.

Biden signed the order at the White House on Monday, requiring agencies to set the standards for that testing and address related cybersecurity, nuclear, radiological, biological, and chemical risks.

The president said that to realize the promise of AI while avoiding the risks, “we need to govern this technology.” He added that in the wrong hands, it could disrupt society through vulnerability exploitation, among other tactics.

The guardrails have prompted mixed responses

Bradley Tusk, the CEO of Tusk Ventures, a venture capital firm invested in AI and tech, welcomed the move. He said tech companies would likely oppose sharing proprietary data with the government over fears it could be given to rivals.

He added that without a “real reinforcement mechanism,” which the executive order does not seem to have, the concept means well, but adherence could be hampered.

NetChoice, a national trade association with influential tech platforms as members, described the order as an “AI Red Tape Wishlist,” which it thinks could get in the way of companies and competitors looking to enter the marketplace and significantly give the federal government more power over American innovation.

The new order goes beyond the voluntary commitments made by AI companies like OpenAI, Alphabet, and Meta Platforms earlier this year to watermark AI-generated content to make the technology safer. The order directs the Commerce Department to develop guidance for content authentication and watermarking to label items generated by AI for the sake of more transparent government communications.

On the issues of IP and AI training

The order includes a directive to evaluate the systems for IP violations. Recently, writers and visual artists have filed multiple lawsuits against tech companies developing AI, accusing them of theft for using their works as training data sets. Tech companies want to argue that US copyright law’s fair use doctrine protects usage.

The Group of Seven industrial countries plans to agree on a code of conduct for companies working on advanced AI systems, according to a G7 document. According to Max Tegmark, the president of tech policy think tank Future of Life Institute, the United States is “already far behind Europe.”

He asserts that policymakers, including those in Congress, must look out for their citizens by passing laws that can tackle threats while ensuring progress. US officials have warned that AI can supercharge the risks of bias and civil rights violations. Biden’s executive order seeks to address that by calling for guidance to landlords, federal contractors, and federal benefits programs to keep AI algorithms from being used to make problematic issues even worse.

Share to...