FBI deputy director says AI helps agents act on threats in hours, not weeks
The FBI has begun using artificial intelligence to determine which threats deserve immediate attention, a move that Deputy Director Christopher Raia says is compressing what used to be weeks of investigative groundwork into hours or minutes.
In an interview with CBS News, Raia described the technology as a practical response to a problem that has long frustrated federal investigators: the sheer volume of tips flowing into the bureau every day. “We’ve sprinkled in AI and now these tips rise to the top,” he said.
At the FBI’s National Threat Operations Center, algorithms now help sort incoming threats, flagging those that may require urgent action before they are passed to field offices. Raia, a 23-year FBI veteran who became deputy director in January, said the system has allowed the bureau to move with much greater speed without asking analysts to work through more data than any human team could manage.
Raia’s interview came three weeks after the bureau released its annual crime report, which showed the U.S. murder rate reaching a 20-year low in 2025 and violent crime overall down by the largest one-year margin on record. Aggravated assaults fell 7.2%, rapes declined 7.6%, and robberies dropped 18.5%.
Raia credited a “combination of things” for the decline, including local police work, what he called the border “being sealed up,” and a renewed focus on violent crime. But when asked about the role of AI, he was blunt about the pressures the FBI has faced in the past. “We didn’t have the manpower to analyze it,” he said, describing a system in which agents were “sitting on a bunch of risk” as hundreds or thousands of leads poured in from around the world.
AI has narrowed that bottleneck, he said, and in some instances it has pushed what might have taken days or weeks to the front of the line. Raia pointed to the case of Zachary Charles Newell, a North Carolina man sentenced in March to two years in federal prison for posting online threats to “shoot up a black preschool” and harm children. In that instance, AI flagged the threat for immediate review and agents acted quickly, he said.
The use of AI to screen tips and online speech has generated questions about whether civil liberties are being protected. Raia said the bureau has kept Congress and outside oversight boards informed and has sent nearly 50,000 pages of documents for review over the past 18 months. “Somebody has to police the police, right?” he said. “We are not taking away the privacies or the civil liberties of Americans, whether it’s AI or any of the other responsibilities that we have as an agency.”
Privacy advocates have also raised concerns about automated license-plate readers, including the network operated by Flock Safety, which says it provides cameras to about 6,000 law-enforcement agencies. Those data feeds have been shared with federal agencies, and Raia said the cameras are important investigative tools, particularly for tracking down leads and suspects. “As far as the cameras go, those are vital in our investigations as far as leads and as far as finding folks,” he said.
Raia also discussed the Trump administration’s push against drug cartels and transnational criminal organizations. The Homeland Security Task Force, established by executive order in January 2025, has put the FBI in closer working relationships with Homeland Security Investigations, the Drug Enforcement Administration, the Treasury Department and the Department of Defense. In August, that task-force initiative led to charges against 81 alleged members of a violent gang in Puerto Rico, accused in a drug-trafficking and firearms-conspiracy case.
Raia described the new systems as a way to get the right tips to agents, not as a substitute for human judgment. The open question, as the FBI applies the technology across more investigations, is whether that distinction will remain clear as pressure grows to respond faster to threats.