Vice President Harris watches as President Biden signs an executive order on artificial intelligence on October 30th. On Thursday, the Biden administration announced new rules for how government agencies deploy AI.
Brendan Smialowski/AFP via Getty Images
hide caption
toggle caption
Brendan Smialowski/AFP via Getty Images
Vice President Harris watches as President Biden signs an executive order on artificial intelligence on October 30th. On Thursday, the Biden administration announced new rules for how government agencies deploy AI.
Brendan Smialowski/AFP via Getty Images
In a memo released by the Office of Management and Budget, the Biden administration announced new guidance for federal agencies on how they can and cannot use artificial intelligence.
This is an important step towards ensuring the safe use of AI, which private companies and other countries are also working on.
The draft guidance was released last fall ahead of Vice President Harris' visit to the first World AI Summit in the UK. The draft was then released for public comment before the final version was released on Thursday.
Harris said the guidelines are “binding” and emphasized the need for guidelines that prioritize the global public interest.
“President Biden and I intend for these domestic policies to be a model for global action,” Harris said on a call with reporters Wednesday. “We continue to call on all countries to follow our lead and prioritize the public interest when it comes to government use of AI.”
Guidance to government agencies strives to strike a balance between managing artificial intelligence risks and fostering innovation.
It also requires each government agency to appoint a chief artificial intelligence officer, a senior role that oversees AI implementation. It also outlines how the government plans to grow its AI-focused workforce, including hiring at least 100 experts in the field by this summer.
“The public deserves to trust that the federal government will use technology responsibly,” said Shalanda Young, director of the Office of Management and Budget (OMB).
Government agencies must adopt AI safeguards by December 1st
Guidance to government agencies states that any AI technology used must have appropriate safeguards in place by December 1st. If you are unable to provide these safeguards, you should cease using the technology unless you can demonstrate that its use is necessary. Agency functions.
Required safeguards include assessing, testing, and monitoring the impact of AI technology, but the details of that process are still unclear in this guidance.
Alex Reeve Givens, president and CEO of the Center for Democracy and Technology, told NPR what the testing requirements are and who in government has the expertise to greenlight the technology. He said he still had doubts.
“I think this is the first step. What comes next is a very detailed practical guide and expectations of what an effective audit looks like, for example,” Reeve Givens said. said. “There's still a lot of work to do.”
One of the next steps Reeve Givens is watching is the guidance the administration releases regarding the procurement process and what requirements it will place on companies with AI technology that it wants to acquire.
“It's really a tipping point where a lot of decisions and values are made and a lot of testing is done before the government puts money into the system in the first place,” she said.
Transparency from government agencies allows for better oversight
Reeve Givens said the part of the guidance from OMB on transparency is particularly noteworthy.
Under the new guidance, agencies will be required to share online an annual inventory of their use of AI and the risks associated with it, and that inventory must be accessible. Reeve Givens said this provision was “key.”
“Then you can ask the question, 'What kind of testing did you do? What did it look like?'” “This provides us with a hook to start a public conversation,” she said.
However, the Department of Defense and intelligence agencies are exempt from sharing information about their use of AI.
This guidance could be a 'catalyst' to encourage further use of AI
OMB guidance also provides for fostering innovation through AI. Ifeoma Ajunwa, a law professor at Emory University, told NPR that the guidance sends a signal to government agencies that it's OK to consider using AI technology.
“I think this is a trigger for agencies that maybe have had some apprehensions or hesitations about using AI technology,” she said.
“I don’t want government agencies to take this as a complete decision to use AI technology in all cases,” Ajunwa added. “But I want them to look at this as an opening and when the safety guardrails are in place they can use them at the appropriate time.”
Some government agencies are already using artificial intelligence, but the Biden administration memo notes that the technology could have an impact on everything from predicting extreme weather to tracking the spread of disease and opioid use. The method is outlined.
