Federal agencies should “commit to deeper, consistent, long-term engagement” in development of AI technology, under new standards issued by the National Institute of Standards and Technology.
The standards come in response to a February executive order instructing that agency to develop a plan to “ensure that technical standards minimize vulnerability to attacks from malicious actors and reflect federal priorities for innovation, public trust, and public confidence in systems that use AI technologies; and develop international standards to promote and protect those priorities.”
According to a summary by the NIST, the plan:
* Recommends the federal government bolster AI standards-related knowledge, leadership and coordination among agencies that develop or use AI; promote focused research on the trustworthiness of AI systems; support and expand public-private partnerships; and engage with international parties.
* Emphasizes the need for agencies to be flexible in selecting AI standards for use in regulatory or procurement actions.
* Calls for prioritizing multidisciplinary research and expanding public-private partnerships to advance reliable, robust and trustworthy AI.
* Highlights related tools that will be needed to support AI, including benchmarks, evaluations and challenges that could drive creative problem solving.
The document is at https://www.nist.gov/news-events.