In a long-running Freedom of Information Act (FOIA) case currently before Judge James Boasberg, on April 15, 2026, the Federal Bureau of Investigation (FBI) admitted in a court filing that, contrary to its May 2, 2019, claim to Cato that it had "no records" on the organization, it had investigated Cato employees and the Institute itself. 

What follows is how we got here and what it means in light of the ongoing congressional debate over whether or not to renew, modify, or kill the Foreign Intelligence Surveillance Act (FISA) Section 702 program and other statutorily unregulated FBI surveillance and investigative practices.

Background

In the early spring of 2019, and as part of a larger FOIA-driven effort to determine the extent of FBI surveillance of domestic civil society groups, I filed a FOIA with the Bureau seeking records mentioning the Cato Institute. On May 2, 2019, the Bureau responded, claiming it had no records on the Institute. Given that Cato scholars 1) routinely take positions at odds with federal law enforcement and intelligence agencies; 2) routinely communicate with persons outside the United States and engage in international travel themselves; and that 3) some Cato scholars reside outside the United States, I assessed that the FBI's "no records" response was not credible and appealed the Bureau's determination to the Justice Department's Office of Information Policy (OIP).

On June 6, 2019, OIP not only affirmed the FBI's "no records" response but also went even further, with then OIP Associate Chief of Administrative Appeals Christina D. Troiani stating:

Furthermore, I have determined that the FBI properly refused to confirm or deny the existence of any national security or foreign intelligence records responsive to your request because the existence or nonexistence of any such responsive records is currently and properly classified. See 5 U.S.C. § 552(b)(1). Please be advised that the Department Review Committee will determine whether the existence or nonexistence of this category of records should continue to be considered a classified fact. Additionally, the existence or nonexistence of any such responsive records is protected under the FOIA pursuant to 5 U.S.C. § 552(b)(3). This provision concerns matters specifically exempted from release by a statute other than the FOIA (in this instance, 50 U.S.C. § 3024(i)(1), which pertains to the National Security Act of 1947 and the Central Intelligence Agency Act of 1949).

Troiani's "refused to confirm or deny" language is what's known in legal parlance as a "Glomar response"—based on a long-infamous DC Circuit decision involving the CIA's then-successful effort to avoid a yes-or-no response to whether it had used Howard Hughes's Glomar Explorer deep-sea mining vessel to recover the sunken Soviet submarine K?129 from the Pacific Ocean floor (it had, but kept that fact secret until 2010). 

The one thing I've learned in studying previous "Glomar response" cases is that the responses are, in fact, de facto acknowledgments that the activity at issue was or is real, and thus the records sought in connection with it do in fact exist. Accordingly, I waited until December 2019, filed a fresh FOIA seeking the same kinds of records, and, after the Bureau failed to produce any, Cato filed suit on November 17, 2020.

In the more than five years since, the Bureau has fought disclosure of records on Cato at every opportunity—but thanks to some key court rulings made possible by great FOIA litigation work by Loevy and Loevy of Chicago, the Bureau has released nearly 1,000 pages of previously "nonexistent" records on the Institute. 

What it has not done is release the most explosive material at issue in this case—the actual investigative records on Cato itself and an as-yet-undisclosed number of its employees—while continuing to invoke law enforcement and intelligence surveillance FOIA exemptions to keep those records secret.