"We're trying to use AI...to make these little tiny decisions, and then bring that up to a human," Mark Nehmer told an audience at the Defense One Tech Summit in Virginia.
Mark Nehmer on AI-assisted vetting
Nehmer, identified by the Defense Counterintelligence and Security Agency as an analytics and innovation chief, said the agency can use artificial intelligence to pare parts of the security-clearance vetting process from "months to hours." Speaking on a panel, he described a workflow in which AI handles narrowly defined tasks that are then elevated to human reviewers, producing a "package of evidence" that a senior analyst can use to reach a final decision.
He did not specify which AI systems would be deployed for that work.
Defense Counterintelligence and Security Agency's role
The Defense Counterintelligence and Security Agency (DCSA) is the Defense Department’s main agency for conducting background investigations and vetting personnel for access to classified information, and the agency serves as a key determinant for whether companies are eligible to work with military and intelligence agencies. DCSA has led the government’s background-check process since 2019, when the Office of Personnel Management handed off its National Background Investigations Bureau to the Pentagon.
Congressionally-approved acquisition overhaul and 43,000 clearance requests
Nehmer tied the push to automate to a recent congressionally-approved acquisition overhaul that encourages defense officials to prioritize goods and services from the commercial market. He estimated that change will require DCSA to process roughly 43,000 clearance requests per year — a scale he framed as one driver for accelerating vetting through automation and AI-assisted decision support.
Trusted Workforce 2.0, continuous vetting, and modernization challenges
DCSA has already enrolled millions of clearance holders in continuous vetting under an initiative known as Trusted Workforce 2.0. But the broader modernization effort, the panel and subsequent reporting note, has not been smooth: it has faced repeated delays, cost overruns and congressional scrutiny. The agency’s pivot toward AI sits atop that longer-running overhaul, adding a new technical layer to reforms that lawmakers and auditors have repeatedly examined.
Export-control action on Anthropic models and the federal AI landscape
Over the weekend, the United States invoked an export-control mechanism that effectively bans two major Anthropic frontier models, a move the reporting says has escalated debates over how Washington can exert control over AI usage in the government. The decision, the source reports, has been widely criticized — a development that arrives as DCSA and other federal entities contemplate expanded uses of AI in sensitive national-security functions.
What this means for senior analysts, cleared companies, and policymakers
- Senior analysts: Nehmer described an operational model in which AI produces "little tiny decisions" and an evidence package that a human senior analyst will use to reach conclusions. Analysts will therefore be watching how those packages are constructed and whether they produce defensible, auditable recommendations.
- Cleared companies and contractors: Because DCSA serves as a gatekeeper for firms seeking to do work for military and intelligence agencies, the agency's ability to process an estimated 43,000 additional clearance requests per year could materially affect who can compete for government business — and how quickly they can be onboarded.
- Policymakers and oversight bodies: The modernization program has already drawn attention for delays, cost overruns and congressional scrutiny. Policymakers interested in accountability will likely press for details on which AI systems are used, how findings are validated, and how the agency will reconcile efficiency gains with legal and security requirements.
DCSA’s public pitch is straightforward: shorten long adjudication timelines by letting machines handle routine steps and humans finalize judgment. The agency’s announcement that AI can reduce parts of the vetting process from "months to hours" is a bold claim; Nehmer offered a snapshot of intent but not the technical or procedural specifics that would allow independent assessment.
That lack of specificity matters because the modernization program that would host such tools has been the subject of repeated delays and costs, and because a recent export-control decision affecting Anthropic models has intensified questions about which systems the government may rely on. The next concrete steps DCSA must take are clear in practice if not yet in public: name the tools, describe the safeguards, and show results that withstand courtroom and congressional scrutiny.
GovExec Editor-in-Chief Frank Konkel contributed to this report.
