Federal AI Efforts Push to Prove Mission Value

Federal agencies are deploying artificial intelligence use cases that are directly tied to mission amid easing regulations as part of the White House’s directives over the past year.

The AI Action Plan calls on the National Institute of Standards & Technology (NIST) to revise its AI Risk Management Framework to ease some of the regulatory burdens, for example.

The revised version is intended to facilitate that and deploy federal AI use cases that “see a payoff that’s commensurate to the investment,” said NIST AI and Cybersecurity Researcher Martin Stanley during GovCIO Media & Research’s AI Summit on Friday in Tysons, Virginia. “It’s very important that people trust their AI, or else they won’t use it. And then that investment becomes not a good investment.”

The value-driven approach to AI development has introduced a higher risk tolerance across government. “The reality is that there’s a calculated risk you have to take,” Head of Federal Sales at Cloudflare Anish Patel said at the event.

This has resonated in efforts like the Pentagon’s November tech acquisition strategy to speed up technology development and delivery.

“An 85% solution in the hands of our armed forces today is infinitely better than an unachievable 100% solution endlessly undergoing testing,” War Secretary Pete Hegseth said in November.

Both speed and trust are key components to services like the Navy, which is building a hybrid fleet of both manned and unmanned platforms operating in a distributed environment.

“We need to operate without necessarily a whole lot of direct bidirectional comms to the beach, and so, to be able to do this in a somewhat autonomous manner,” said Chris Page, deputy director of the Chief of Naval Operations’ intelligence division. “There’s a point, the earliest time intelligence of value and the latest time intelligence of value, and we need to make sure that our analytical processes … deliver the answer that’s needed within the right time frame. And there’s a level of accuracy required.”

The Marine Corps stood up three digital transformation teams over the past year to deliver emerging technologies to the command and integrate into Marine Corps headquarters to inform policies and standards, said Capt. Christopher Clark, AI lead for the service’s Deputy Commandant for Information. The service is currently developing additional teams to launch over the coming year.

“The reward has to be significant to invest in some of this technology. It is very expensive, and so having those teams out there where they’re embedded in the mission problem, they understand the problems they face day-to-day, and bringing that information back to headquarters Marine Corps, where we aren’t necessarily in that fight, is extremely valuable in understanding where is the value proposition, what is going to make the Marines faster, more lethal, make better decisions,” Clark added.