The Wrap: NIST Eyeing AI Black Boxes; FedRAMP Launches Digital Pilot; Space-Age Deterrence

Welcome to The Wrap for Thursday, August 29!

From the newsroom at MeriTalk, it’s the quickest read in Federal tech news. Here’s what you need to know today:

?

NIST Eyeing AI Black Boxes

Let’s face it – few of us normies are ever going to be able to tightly grasp the precise workings of ever-advancing artificial intelligence (AI) models, so it’s a good idea that the U.S. Artificial Intelligence Safety Institute (AISI) is taking on that job with the mission of advancing AI safety. AISI – which is housed within the National Institute of Standards and Technology (NIST) and owes its creation to the Biden administration’s 2023 AI executive order – has signed onto first-of-their-kind research and testing agreements with Anthropic and OpenAI to take a look inside what are black boxes to most and assess how safe they are to use before they hit the commercial market. The agreements, AISI said, will enable collaborative research on AI capabilities and safety risks, as well as ways to mitigate those risks. “Safety is essential to fueling breakthrough technological innovation. With these agreements in place, we look forward to beginning our technical collaborations with Anthropic and OpenAI to advance the science of AI safety,” Elizabeth Kelly, the director of the U.S. AISI, said in a press release today. “These agreements are just the start, but they are an important milestone as we work to help responsibly steward the future of AI,” she said.

?

FedRAMP Launches Digital Pilot

The General Services Administration’s ( GSA ) Federal Risk and Authorization Management Program (FedRAMP) wants industry to get on board with its new Digital Authorization Package pilot that the program unveiled this week. The pilot will explore use of the Open Security Controls Assessment Language (OSCAL) to create machine-readable, digital authorization packages – a longtime goal of the program which is undergoing an extensive operational overhaul mandated by legislation approved by Congress in late 2022. The FedRAMP team is looking for collaboration partners, including cloud service providers (CSPs); governance, risk, and compliance (GRC) tool providers; and Federal agencies to review and use new FedRAMP open source guidance and validation tooling, and then share feedback with FedRAMP. Speaking today at an Information Technology Industry (ITI) Council event, Federal Chief Information Officer (CIO) Clare Martorana pounded home the message that GSA is craving public input. “We don’t do public comment for sport. We don’t do public comment because we have to. We actually don’t need to public comment,” Martorana said. “We’ve tried to approach technology in this administration differently,” she said, adding, “we read every single public comment that we get.”

?

Space-Age Deterrence

As the space realm continues its rapid growth for all kinds of civilian and military applications – along with the potential for weaponization to defeat them – a recent Pentagon senior leadership alumni put into perspective this week how important it is to create credible deterrence against potential attacks. Speaking at GovExec ’s inaugural Space Project event in Washington on Aug. 28, former Assistant Secretary of Defense for Space Policy John Plumb talked about the value of being able to react very quickly to military mischief in space. “Building an infrastructure that allows you to quickly respond, not take two, three, four, five years to build a rocket, build a payload, but just in a month or a few days, ideally, and be able to show that you can launch into a design orbit for that type of mission is a new thing,” Plumb said. “It’s pretty interesting. I think it’s actually pretty important for deterrence, frankly, and to get to the point of deterrence to be credible.” The United States Department of Defense , in partnership with the Defense Innovation Unit (DIU) , announced last August that it’s expanding the scope of its upcoming mission to launch a satellite on 24 hours’ notice and hopes to reach an initial operational capability for “Tactically Responsive Space” in 2025.

?

Intel’s AI Enthusiasm, Caution

On the intelligence agency side of life, AI tech as a double-edged sword continues to be at the top of the list for current agency leaders who spoke at the AFCEA/INSA Intelligence & National Security Summit on Aug. 28. Illustrating the numerous facets of thought on the subject were Central Intelligence Agency Deputy Director David S. Cohen who explained that AI has the capability to assist analysts in sifting through mountains of data to create useful situational awareness, and Gen. Timothy D. Haugh, who heads the National Security Agency and U.S. Cyber Command , who said that AI is increasingly helping analysts focus “on the right piece of information.” Voicing more cautious tones were Federal Bureau of Investigation (FBI) Deputy Director Paul Abbate, who explained that AI is also letting adversaries up their games, and Frank Whitworth, director of the National Geospatial-Intelligence Agency (NGA), who warned against viewing AI through rose-colored glasses. He called for a pragmatic approach, urging caution against overly optimistic expectations and emphasizing the need for practical, grounded strategies. “Idealists might have you think it’s going to be perfect,” said Whitworth.

?

Once again, let’s “call IT a day,” but we'll bring you more tomorrow. Until then please check the MeriTalk breaking news website throughout the day for the latest on government IT people, process, and policy. And finally, please hit the news tip jar [with leads, breaking news, or simply your two cents] at [email protected].