DISA Director Says Data Issues Causing DOD to Lag Behind in AI Implementation

The technical director of the Defense Information Systems Agency’s Cyber Development Directorate said the Department of Defense continues to lag behind the private sector in implementing artificial intelligence for cybersecurity purposes.

Speaking at an Advanced Technology Academic Research Center event, Drew Malloy explained that the DOD is having a hard time ensuring the validity and cleanliness of data, making it difficult for the department to implement AI. He said the Pentagon needs to look at the end-to-end supply chain for an AI model to understand the technology and associated data better.

Malloy’s comments are in line with a June Government Accountability Office report that urged the Pentagon to create department-wide guidance on AI technology acquisitions, Nextgov/FCW reported.

According to the GAO report, the DOD should consider private company practices as a model for developing its guidance. The oversight body solicited information from 13 agencies about how they acquire AI capabilities, all of which include understanding mission needs, testing and evaluating solutions and planning for future efforts.

GAO also pointed out that the Pentagon’s historical struggle with acquiring AI poses additional challenges to other acquisition programs, such as weapons systems.