Today in Security - U.S. Standards for Artificial Intelligence

Today in Security: U.S. Standards for Artificial Intelligence

​The U.S. Department of Commerce's National Institute of Standards and Technology (NIST) has issued a draft plan for federal government engagement in advancing artificial intelligence (AI) standards for U.S. economic and national security needs.

NIST is now seeking public comment on the draft plan, U.S. Leadership in AI: Plan for Federal Engagement in Developing Technical Standards and Related Tools, the agency announced on Tuesday.

The draft describes four levels of possible engagement by the federal government: monitoring, participating in, influencing, or leading standards efforts. It also states that the government should prioritize efforts that are inclusive and accessible, open and transparent, consensus-based, globally relevant and nondiscriminatory, and that use multiple approaches.

The document recommends four actions: promote focused research on the trustworthiness of AI; bolster AI standards-related knowledge, leadership, and coordination among federal agencies; support and expand public-private partnerships; and engage with international parties.

According to the draft plan, trustworthy AI include accuracy, resiliency, safety, security, and the ability to protect privacy. The draft says that increasing trust in AI is a key element in accelerating adoption of AI technologies for economic growth and future innovations.

"AI is already transfor​ming so many aspects of our lives and has the potential to do much more," NIST Director Walter G. Copan said in a statement. "The federal government has an important role to play, in partnership with industry and academia, to ensure that the U.S. maintains its leadership in AI.

Federal officials said the draft plan as a step forward in implementing the American AI Initiative, a national strategy to strengthen America's leadership in AI. Most experts say the United States is battling China for AI dominance, and many say that while the United States may lead in innovation, China is leading in implementation.​​

The draft argues that the widespread use of standards will facilitate technology advancement by defining and establishing common foundations for innovation. The draft plan also outlines the current state of AI standardization, listing categories in which technical standards are needed, including data, metrics, safety, and trustworthiness.

"We want to make the most of this technology, while ensuring safety, privacy and security," Copan said.

Public comment on the draft plan may be submitted to [email protected] through 19 July 2019.  A final plan is due to the White House on 10 August 2019.

For a brief discussion of how AI could be used for rogue purposes, see the Security Management article "Artificial Adversaries."​