Having adopted the Defense Innovation Board’s (DIB) ethical AI principles last February, a senior official with the Department of Defense’s (DoD) Joint Artificial Intelligence Center (JAIC) provided an update today on how the department is approaching those principles a year later.

Alka Patel, who leads JAIC’s ethical AI adoption strategy, explained at today ACT-IAC Emerging Technology and Innovation event how DoD has been adjusting it’s thinking on AI over the past year and operationalizing the principles of the AI ethics strategy.

“We look at the people aspect, we look at organizational culture, and operating structures collectively and holistically to really try to put this into practice,” said Patel. “And I will say too that a lot of this, again, is not something we can solve alone … This is really a conversation where we want to include our industry partners and our academic partners. There’s still a lot to learn.”

A three-part virtual event series showcasing real technology solutions making an impact today. Learn More

Patel broke down the five principles of AI ethics, which are: responsible, equitable, traceable, reliable, and governable. She said it’s important to ensure that the workforce has an appropriate understanding of the technology, data, processes, and development, when it comes to traceability, reliability, and governability. For equitable and responsible, Patel said that JAIC is focusing on ensuring that data sets are representative, and being aware of data bias concerns.

Additionally, Patel discussed the need for having interoperability from a cultural perspective – not just a technology perspective.

“One of the areas that we’ve been really thinking a lot about is in the acquisition space and as we think about these technologies – we’re procuring these technologies – and so these industry partners have perhaps their own principles or perhaps their own processes,” said Patel.

“And so oftentimes we have this conversation around interoperability from a technical perspective, but I think there’s also a need for a conversation around interoperability from a cultural perspective, from a process perspective, from a principals perspective, and then how do we align along those conversations or along those areas so that we engender … trust,” she said. “This is such a complex conversation for sure.”

Read More About
More Topics
Jordan Smith
Jordan Smith
Jordan Smith is a MeriTalk Senior Technology Reporter covering the intersection of government and technology.