Photo: Shutterstock

The U.S. House of Representatives' Oversight and Government Reform Subcommittee on Information Technology released a white paper this month calling for the U.S. government's restraint in passing new regulations around artificial intelligence (AI) technology, and more investment for the development of AI.

The 16-page white paper, titled “Rise of the Machines: Artificial Intelligence and its Growing Impact on U.S. Policy,” said the federal government should approach creating new regulations by first assessing if AI technology's risks to public safety are already addressed by existing laws. If those risks fall outside existing regulations, the federal government “should consider carefully whether modifications or additions are needed to better account for the addition of AI.”

The hesitancy in drafting new regulations, according to the paper, stems from wanting the United States to become a leader in AI by promoting private sector innovations.

However, at a minimum, the subcommittee wrote that there should be a “widely agreed upon standard” for measuring AI products' safety and security before any new regulations are enacted.

The National Institute of Standards and Technology (NIST) was cited as “a key player in developing standards,” along with the Institute of Electrical and Electronics Engineers' Global Initiative on Ethics of Autonomous and Intelligent Systems. The subcommittee said the federal government should support public, educational and private sector efforts in developing standards for measuring the safety and security of AI products.

Marc Canellas, a voting member of the IEEE-USA Artificial Intelligence and Autonomous Systems Policy Committee, welcomed the subcommittees advice, but wanted the government to take a more active role in regulating AI regulations.

“As much as people don't talk about this, sometimes having standards about what you can and cannot do is helpful,” Canellas said.

He added that he would like to see more action taken outside of just issuing reports. “My main issues with this, especially with working on the Hill, is Congress is super good at writing reports and having hearings and panels. What their job is is to introduce legislation. … Where are the bills?”

Canellas said that to test the federal government's ability to create rules that do not negatively affect commercial industry, AI regulations can first be implemented in the criminal justice system, the National Security Agency and the U.S. Department of Transportation.

The white paper stressed the importance of the United States being a leader in artificial intelligence development. The U.S. no longer reigning as a leader in AI, it argued, could pose a risk to ensuring nation-states abide by international humanitarian laws concerning AI warfare. It may also affect the country's economic growth to make industries more efficient.

Still, the subcommittee's white paper also highlighted the problems AI can pose to the public. McKinsey Global Institute and Oxford University reports cited by the white paper found that AI advancement could lead to short-term job loss from the automation of some jobs, though other witnesses and studies provided evidence that AI could improve and increase jobs. A recent PwC study also concluded that artificial intelligence could create more job hires than cuts in the legal profession.

The white paper further noted that the vast amount of personal data AI relies upon and stores poses a privacy challenge, while AI could also be used in making “'fake news'” seem more convincing and allow autonomous vehicles to be “tricked” by hackers. 

What's more, biased or skewed data used by AI programs in criminal law agencies and courts may have long-standing negative consequences for individuals, the report stated. “There are legitimate concerns that if an AI system is trained on biased data, the AI system will produce biased results.”