When Health Secretary Robert F. Kennedy Jr. defended his vision of the Centers for Disease Control and Prevention in a recent op-ed, he cited one of the agency’s biosurveillance programs as a prime example of the agency’s capabilities: the Biothreat Radar Detection System.
But the “Biothreat Radar Detection System” doesn’t appear to exist — at least, not yet, sources inside and outside the CDC told NOTUS. And new details about how the program might apply AI to biosurveillance are giving biosecurity experts some pause.
“We have shown what a focused CDC can achieve,” Kennedy wrote in a Wall Street Journal opinion piece titled “We’re Restoring Public Trust in the CDC.” “Its Biothreat Radar Detection System — an advanced early-detection tool — can spot pathogens like H5N1 or MERS early enough to prevent catastrophe.”
NOTUS was unable to find evidence that the Biothreat Radar Detection Program is currently receiving any funding, though the administration’s 2026 Health and Human Services and CDC budgets propose allocating $52 million to launch the program.
One CDC employee who was granted anonymity to speak candidly about their employer said that while their colleagues within one of the agency’s biosurveillance programs were aware of a Biothreat Radar program proposal, none of them knew of any ongoing work happening to support the initiative.
Biosecurity experts outside the agency with deep knowledge of the U.S.’s biosurveillance capabilities were also stumped by Kennedy’s reference to the program in his op-ed.
“I literally wondered what he was talking about,” Gigi Gronvall, a professor at the Johns Hopkins Bloomberg School of Public Health, said via text.
Kennedy himself was seemingly aware of the fact that the Biothreat Radar program wasn’t up and running only days before the Wall Street Journal op-ed was published.
In an August 28 email to CDC employees that was reviewed by NOTUS, Kennedy wrote, “We will soon launch the Biothreat Radar Detection System to detect threats like H5N1 and MERS before they spread, strengthening our defenses against future pandemics.”
An HHS spokesperson, Emily Hilliard, didn’t respond to NOTUS’ questions about whether the Biothreat Radar Detection System is currently operational but described it as “a project designed to combine new technological capability with automated result interpretation by artificial intelligence systems.”
Hilliard referred to it as a “proposed” system and said it would be able to screen groups of people for infectious diseases, not just individuals, by combining “cutting edge American technologies and the development of advanced AI hardware and algorithms” to “replace and expand” the CDC’s current infectious disease surveillance programs.
“Early systems for this exist and will serve as the base for an expanded effort,” Hilliard wrote, “but the digitalized data will be subjected to higher-order processing to enhance sensitivity and automatic warning on a large national public health scale.”
Biosecurity experts told NOTUS that while an AI-powered biosurveillance system would be a new — and potentially helpful — addition to the CDC, they cautioned that it could come with technological and ethical quandaries.
Using AI to combine biosurveillance data with other information about groups of people could lead to ineffective and potentially harmful generalizations about who is likely to carry a disease, said Jennifer Nuzzo, Director of the Pandemic Center at the Brown University School of Public Health.
“There’s peril involved in that,” Nuzzo said. “It used to be that you couldn’t immigrate to the United States if you had HIV. Those rules were changed about 10 years ago, but you could imagine the same kind of tendencies that created that rule in the first place being misapplied.”
She added that getting the CDC involved in issues like border control could lead to countries not wanting to collaborate with the CDC because “now it’s a national security engagement, it’s not a health engagement.”
Gronvall pointed out that an AI-powered analysis pipeline would require comprehensive data — so the Trump administration’s cuts to the CDC’s surveillance programs could keep such a program from being effective, she said.
“AI can only give you an accurate analysis if you put real data into it to analyze,” Gronvall said. “The drastic cuts in actual surveillance that will affect all levels — international, states, local — will leave AI with nothing but hallucinations.”
The Biothreat Radar program is in the administration’s proposed 2026 HHS budget and the proposed 2026 CDC budget, neither of which has been finalized by Congress.
Hilliard said the system would feed “diverse generated data streams for pathogen surveillance” into the One CDC Data Platform, an existing data sharing system, using an “AI improved” data integration program to build upon the CDC’s Response Ready Enterprise Data Integration system.
“It is envisioned that an AI-upgraded RREDI system will provide the link to pandemic/biological readiness, emergency readiness and response authorities via its automated data management processes to provide timely insights for response-related decisions,” Hilliard wrote.
Hilliard didn’t give specifics on the exact tools the Biothreat Radar program would use to generate this data. Neither do the proposed budgets. The president’s draft HHS budget describes the initiative as “pathogen agnostic,” meaning it could theoretically test for the presence of any pathogen rather than for specific diseases, like COVID-19 or H5N1, as is commonly done by current CDC biosurveillance programs.
That’s an important distinction because it would allow the CDC to scan for a wide range of pathogens rather than having to know specifically which ones to look for, said Jeff Kaufman, director of the Nucleic Acid Observatory, part of a Boston-based nonprofit that’s trying to develop a pathogen early detection system. He said such an approach could speed up the detection of a potentially pandemic-causing illness before it spreads enough to cause widespread alarm.
An internal CDC slide on the presidential 2026 budget proposal that was reviewed by NOTUS only briefly mentioned AI. It said the program would obtain samples from “travelers, wastewater, other systems,” via the already-existing Traveler Genomic Surveillance program, the slideshow continued, use AI to “transform data into actionable public health intelligence” and then work with “DOD, state, local and global public health partners” to “address and stop disease threats.”
After reviewing the CDC slide, Kaufman said the broad strokes of the proposal seemed “generally very good.”
Still, he and others doubted whether all of the promises made by Kennedy and the Trump administration were realistic.
The HHS budget proposal says the new program would be able to “rapidly detect novel pathogens with 24-hour turnaround times.” Kaufman said that turnaround time, with the logistics laid out by HHS, was “not realistic.”
“I think if you really optimize your logistics, you can get to a turnaround time of just over two days,” Kaufman said. “In this case, it does involve things like paying people a bit more so you can have a night shift in your wet lab, but I think the efficacy benefits of that are really worth it.”
The ongoing crisis at the agency would also hamper any expanded surveillance plans due to personnel and funding limitations, said Nuzzo.
“If they don’t have the partnerships on the ground to get specimens, there’s nothing to test,” Nuzzo said.
The CDC has lost thousands of employees since Trump took office earlier this year. Many of those personnel were let go by DOGE. Some took voluntary buyouts. Four high-level officials resigned after Kennedy dismissed CDC director Susan Monarez last month.
Nuzzo also said simply spotting a pathogen isn’t the same thing as being capable of effectively responding to a potential outbreak.
“Just detecting a pathogen and characterizing what that pathogen is and modeling it does not equal preparedness,” Nuzzo said. “Preparedness means trying to understand the context in which the specimens were generated, being on the ground in order to intervene, to do further investigation. It means having relationships with partners overseas to understand what’s going on, and to talk about what efforts need to be done.”
“There are deep questions about CDC’s ability to do that,” Nuzzo said.