Mariam Baksh

Key federal officials including the government's top information security officer said they are raring to adopt artificial intelligence for applications like cybersecurity, during appearances at a major conference, as representatives from industry and academia urged greater efforts to understand and monitor the technology.

The Bureau of Industry and Security has tripled a key metric, as identified in President Biden's executive order -- EO 14110 -- on artificial intelligence, for determining covered entities in a rule the agency has proposed for requiring developers and potential developers of dual-use AI foundation models to report “applicable activities” to them once every three months.

Artificial intelligence scholars from various institutions are highlighting the National Institute of Standards and Technology's failure to account for “marginal risk” in drafting guidance for developers to manage the threat of foundational AI models being misused.

The House Science Committee has passed major legislation to establish the National Artificial Intelligence Research Resource, codify a red-teaming consortium to help the National Institute of Standards and Technology develop evaluation standards for AI, and require NIST to issue more related guidance.

The National Artificial Intelligence Research Resource is crucial for leveling the playing field so public and smaller entities can keep pace with the mega companies developing the technology, according to a digital rights advocate who highlighted a need for transparency as legislation to codify the effort advanced out of the House Science Committee.

The Biden administration's landmark executive order on artificial intelligence employs a flawed strategy for evaluating the risks presented by “dual-use” AI foundation models, Alan Davidson, head of the National Telecommunications and Information Administration, told an audience of his former peers in the scientific community.

A senior advisor to the Center for Democracy and Technology, commenting on Nevada's contract with Google to more quickly process the state's backed-up appeals for unemployment benefits, said the development presents an actual danger of using the technology.

Draft guidance the National Institute of Standards and Technology has issued to help developers of dual-use AI foundation models mitigate risks of the technology being misused creates confusion around the agency's consideration of elements like bias and discrimination, HackerOne said, amid a chorus of stakeholder feedback on the document.

The National Institute of Standards and Technology’s draft guidance to help developers of dual-use AI foundation models mitigate risks of the technology being misused creates confusion around the agency’s approach to consider elements like bias and discrimination, according to HackerOne.

The Biden administration is highlighting efforts to facilitate the development of frontier artificial intelligence in the United States, including by significantly reducing the time it takes to conduct environmental impact studies for permitting the energy- and water-intensive data centers associated with the technology.