In less than 25 days, the legislative authority for America’s most important foreign intelligence collection program will lapse. Section 702 of the Foreign Intelligence Surveillance Act (FISA) will expire on April 20, 2026 unless Congress acts to renew it. In the middle of a war, in a deeply polarized and contentious world where the United States faces a range of adversarial threats, the security of the nation should never face the prospect of such a serious, avoidable lapse. But, because Section 702 is not a permanent part of FISA, it must receive periodic reauthorization from Congress and its renewal is at risk today. During the last reauthorization debate, which ended with Section 702 continuing under the mandates of the Reforming Intelligence and Securing America Act (RISAA) in April 2024, the President’s Intelligence Advisory Board (PIAB) observed: “The cost of failure is real. If Congress fails to reauthorize Section 702, history may judge the lapse of Section 702 authorities as one of the worst intelligence failures of our time.” And, yet, here we are again.
The critical value that the Section 702 intelligence collection program contributes to the nation’s security is beyond dispute. It is, of course, the nation’s most valuable intelligence tool in combatting terrorism and foiling malicious cyber activity, two of Iran’s favorite asymmetric warfare specialties, for example. During the last reauthorization debate in the 2023-2024 cycle, the government reported that this single collection asset supported 60% of the articles in the President’s Daily Brief; that 70% of the CIA’s illicit synthetic drug disruptions with foreign partners stemmed from Section 702 data; that 70% of the CIA’s successful weapons and counterproliferation disruption derived from Section 702 data; and, that 100% of the president’s intelligence priorities topics reported on by NSA were supported by Section 702-derived intelligence.
The Section 702 program has been controversial since it was first enacted in 2008, and Congress required that the program be subjected periodically to legislative review and reauthorization. This is the fourth time that it has faced possible termination (Section 702 was previously reauthorized in 2012, 2017, and 2024), and the accompanying debates have remained contentious with a coalition of privacy advocacy groups frequently criticizing the program and calling for its reform.
The most provocative and recurring grievance lodged by these groups is the contention that Section 702 permits warrantless “backdoor” searches of Americans’ electronic communications in violation of the Fourth Amendment; they contend that agencies with access to all, or some part of, the database of communications collected under Section 702’s authority (i.e., NSA, FBI, CIA, and NCTC) are operating an illegal warrantless spying program on Americans’ private information. I have rebutted these allegations in numerous articles over recent years (some of which are here, here, here, here, and here). Recently, the Brennan Center released a piece titled, “FISA Section 702 Backdoor Searches: Myths and Facts” that I argue misstates the law and record governing the querying process. The Center’s overriding error is the false premise that querying the Section 702 database using U.S. Person (USP) identifiers to locate and retrieve foreign intelligence information is a “backdoor” search that violates the Fourth Amendment.
It does not. Let me explain why.
How Section 702 Works – Incidental collection and minimization
Section 702 authorizes the government to target the communications of foreigners reasonably believed to be located outside the United States but, simultaneously, forbids targeting any U.S. person, whether or not they are located in the United States. Congress has always recognized that “it is simply not possible to collect intelligence on the communications of a party of interest without also collecting information about the people with whom, or about whom, that party communicates including, in some cases, non-targeted U.S. persons.” Congress acted to protect the privacy rights of those “non-targeted U.S. persons” subject to this “incidental collection” by requiring that the Attorney General adopt, and the Foreign Intelligence Surveillance Court (FISC) review and approve, procedures that minimize the acquisition, retention, and dissemination of information concerning nonconsenting U.S. persons.
These minimization procedures, along with targeting and querying procedures (the latter representing a requirement Congress added as part of its 2017 reauthorization of Section 702) comprise the statutory architecture that Congress created to protect the Fourth Amendment rights of those U.S. persons whose communications are incidentally collected during lawful Section 702 acquisitions (such as U.S. persons who are communicating with foreign targets). The targeting, minimization, and querying procedures that Congress mandates for every Section 702 acquisition must be reviewed and determined by the FISC (comprised of already serving federal district judges appointed by the Chief Justice) to be consistent with the Fourth Amendment’s requirements.
First, consider the acquisition of the information on the front end of the Section 702 program. “Reasonableness” is the “touchstone” of the Fourth Amendment and that Amendment offers no guarantee that a warrant will be an essential prerequisite to a government search or seizure that might impact individual privacy interests where the search is intended, and reasonably designed, to acquire foreign intelligence information. The requirement of a warrant in a law enforcement context is viewed as a proxy for reasonableness where an individual’s life, liberty or property is at stake; but, where the purpose of the “search” is to acquire foreign intelligence information, not to target or pursue any particular American citizen, courts have repeatedly concluded that a warrant fails to properly balance the interests at stake when the government is not enforcing the law but, rather, is seeking to preserve and protect the national security.
Section 702 acquisitions are initiated by the Attorney General and the Director of National Intelligence (DNI) filing a written Certification with the FISC. Once the FISC approves a Section 702 certification, the Attorney General and the DNI issue one or more “directives” to electronic communication service providers (e.g., Verizon, AT&T), ordering that they furnish the government those communications to or from a list of tasked selectors (e.g., email addresses or telephone numbers) used by the targets of the authorized acquisitions. Communications responsive to the identified selectors are provided to NSA, which is principally responsible for the operational implementation of Section 702 and the keeper of the database of communications acquired pursuant to Section 702 surveillance.
The communications of U.S. persons corresponding with a Section 702 target using a tasked selector will be “incidentally” collected as part of the authorized acquisition of the target’s communications. The scope of “incidental” collection is not insignificant because Section 702 is a large, programmatic surveillance program annually collecting hundreds of millions of communications from its 291,824 foreign targets. Therefore, the total (though not relative) number of “incidentally” acquired U.S. person communications is sizeable. It is this “incidental” collection of U.S. person communications and the subsequent handling of those communications that lies at the heart of the allegations of “warrantless surveillance” and “domestic spying.”
The Proper Fourth Amendment Analysis of Electronic Surveillance Conducted for Foreign Intelligence Purposes
The FISC, the Foreign Intelligence Surveillance Court of Review (FISCR), and other federal appeals courts to consider the issue all have recognized that surveillance conducted for foreign intelligence purposes can be analogized or likened to the Supreme Court’s “special needs” cases (special in the sense of not for law enforcement purposes). That understanding led the FISCR to conclude that a foreign intelligence exception applies to the Fourth Amendment’s warrant requirement because (a) the national security purpose behind the surveillance transcends ordinary law enforcement purpose, and (b) there is a high degree of probability that a warrant requirement would hinder the government’s ability to collect time-sensitive information impeding vital national security interests. For nearly two decades, the FISC (and, when called upon, the FISCR) have employed a “totality of the circumstances” approach recognizing that the correct Fourth Amendment analysis of electronic surveillance conducted for foreign intelligence purposes examines the programmatic purpose served by that surveillance, whether that purpose serves a legitimate national security objective rather than law enforcement, and whether that purpose would be “frustrated” by insisting upon a warrant.
The Querying Process
Communications lawfully collected pursuant to an approved Section 702 acquisition are stored in the Section 702 database retained by NSA where they reside until extracted by means of a query by an analyst. These queries are undertaken consistent with the querying procedures mandated by Congress that each agency having access to any part of the Section 702 database (i.e., NSA, FBI, CIA and NCTC) must develop, and which must be approved by the FISC as consistent with the Fourth Amendment, both as written and as implemented. Each query is comprised of one or more search terms which must be factually based upon a reasonable belief that it will, and be reasonably structured to, find and extract foreign intelligence information.
Querying practices, and more particularly the FBI’s querying practices, have long been the focus of Section 702’s critics’ claim of warrantless “backdoor” searches. Since there is no strong or credible legal basis for challenging the acquisition of communications pursuant to a FISC-approved Certification, the critics’ theory is rather that the Fourth Amendment forbids using a U.S. Person (USP) identifier to extract foreign intelligence information from lawfully acquired communications without a warrant. In other words, these critics advance the view, despite all the federal court opinions described above, that the government must get a court order to examine communications it already has lawfully acquired and stored in its own database (more below, on an outlier federal court decision).
The particular focus of the critics has been the FBI and, it must be acknowledged, the FBI’s past record of compliance with its own querying standards has been a problem. But it’s a problem that the Bureau and intelligence community have acknowledged and repaired. Indeed, starting in the summer of 2021, in the course of reporting noncompliance issues for review by the FISC, the FBI initiated a series of significant remedial measures designed to improve FBI querying practices; with additional systemic and policy changes following in 2022 and 2023. Those changes have resulted in a dramatic decrease of over 99% in the FBI’s total number of USP queries from 2.9 million in 2021 to 5,518 in 2024 with an uptick to 7,413 in 2025. Following implementation of these reforms, both the National Security Division in the Department of Justice (which is tasked with principal responsibility for reviewing FBI querying practices) and the FISC concluded that the FBI’s querying compliance had improved.
Improvements in FBI Querying Practices
RISAA created a series of required reforms for FBI querying practices and
mandated that the Department of Justice Office of the Inspector General (DOJ OIG) conduct an audit to ensure compliance with those reforms. In October 2025, the DOJ OIG issued its report concluding that “the FBI has implemented all of RISAA’s querying reforms” and had taken “concrete steps” to implement those reforms. As a result, the number of non-compliant FBI queries has been substantially reduced post-RISAA. While cautioning that its review covered only one-year post-RISAA, the report noted that the number of noncompliant FBI queries has been “significantly reduced” and that the DOJ OIG was “encouraged” that most of the noncompliant queries post-RISAA were attributable to “administrative mistakes, such as typographical errors, rather than due to fundamental misunderstandings of the query standard.”
A Critical Look at the Brennan Center’s “FISA Section 702 Backdoor Searches: Myths and Facts”
The Brennan Center recently released FISA Section 702 Backdoor Searches: Myths and Facts – a discussion that I contend mischaracterizes both the law governing the querying process and the record of compliance under Section 702. Each claim is addressed below.
As a legal backdrop, the Brennan Center’s central premise — that a warrant is constitutionally required to query the Section 702 database for USP information — has been repeatedly rejected by the FISC, and was rejected by Congress itself in 2017 when the Chairman of the House Intelligence noted that a new querying procedure mandated by statute “does not reflect the committee’s belief or intent that law enforcement access to lawfully acquired information constitutes a separate search under the Fourth Amendment.” (House Intelligence Committee Chairman Chris Stewart (R-Utah). He explained further, “The Fourth Amendment, as interpreted by numerous Federal courts, does not require the FBI to obtain a separate order from the FISC to review lawfully acquired 702 information.” .
Brennan Center: “The FBI conducted over 57,000 “backdoor searches” in 2023 alone.”
Misrepresentation/False. If the Brennan Center’s use of “searches” is suggesting that the FBI’s 57,000 queries using USP identifiers is a Fourth Amendment event that required a warrant, this is incorrect. No warrant is required, and the FISC has repeatedly rejected such a characterization. The Brennan Center says 2023 is “the last year for which the government might have provided complete data,” but that obscures the matter. Indeed, the Brennan Center links to the ODNI report that provides data for 2024. While the Brennan Center uses the 2023 statistics, it should be noted that, in 2024, the year in which the FBI began implementing the mandates of RISAA, the FBI made a substantially fewer 5,518 queries using USP identifiers. That figure – of 5,518 is from the same linked ODNI report. Furthermore, the statistics for 2025 were also reported (a few days before the Brennan Center publication), and the 2025 totals are in the neighborhood of the 2024 total, not the pre-RISAA ones. For 2025, the total came to 7,413.
(Note: The Brennan Center may be excluding 2024 and 2025 on the theory those figures are unreliable; I discuss their reliability below.)
Brennan Center: Congress should close the “backdoor search loophole” by requiring the government to obtain a warrant or FISA Title I order.
No. Congress did not create such a requirement in 2012, or in 2017, or in 2024 when it reauthorized Section 702, and it should not do so now. The Fourth Amendment does not require a warrant and, as the President’s Intelligence Advisory Board cautioned in 2023, “A requirement that an intelligence agency should obtain a warrant or court order prior to every U.S. person query of Section 702-acquired information would prevent intelligence agencies from discovering threats to the homeland.”
The practical consequence of the demand for a warrant requirement for these “backdoor searches” is that it would require the government to obtain a warrant to search its own database to retrieve communications it already has lawfully collected.
Brennan Center: It is a myth that warrantless backdoor searches are constitutional.
No. Rather, it is an accurate statement of the law because the Fourth Amendment’s “reasonableness” standard does not require a warrant to obtain foreign intelligence information from communications already lawfully acquired by the government (see discussion above of FISCR, FISC, and other federal court cases).
Brennan Center: A federal court recently ruled that the Fourth Amendment’s warrant requirement applies to backdoor searches.
Ignores overwhelming, contrary precedent. The ruling referenced by the Brennan Center was issued in January 2025 in a muddled, disjointed opinion from a district court apparently struggling with the intricacies of the Section 702 collection program for the first time. That decision is examined at considerably greater length here. It is a flawed decision that has not been adopted by another court and has been repeatedly rejected by the FISC since its release. Simply put, one district court’s (faulty) ruling cannot displace nearly two decades of review by the FISC and FISCR finding that Section 702, as written and implemented, satisfies the Fourth Amendment’s requirement of reasonableness without a warrant.
The Brennan Center’s document is admittedly brief and to the point, but it would have been better for them to note the contrary case law, and how much this singular district court decision is an outlier.
Brennan Center: It is a myth that any American caught up in Section 702 surveillance is a terrorist.
Yes, this is correct. It’s a myth. The government has never contended otherwise. Section 702 collection is authorized for specific foreign intelligence purposes. That an American’s communications are incidentally collected because they correspond with a foreign target says nothing about the American’s activities or associations.
Brennan Center: It Is a “myth” that the Reforming Intelligence and Securing America Act (2024) significantly reduced the number of U.S. person queries conducted by the FBI and put an end to violations of the rules governing queries.
No, it’s no myth. The historical record refutes this claim. RISAA’s mandated reforms and the FBI’s remedial measures have produced significant improvements in FBI querying practices. As noted above, a thorough review by DOJ OIG, as required by RISAA, concluded in October 2024: “the reforms the FBI has made to its querying practices, including the reforms imposed by RISAA, have significantly reduced the number of noncompliant queries.” The Inspector General was also “encouraged” that most of the noncompliant queries post-RISAA were attributable to “to administrative mistakes, such as typographical errors, rather than due to fundamental misunderstandings of the query standard.” Equally significantly, as noted earlier, after RISAA, the FBI reduced the number of U.S. queries from more than 57,000 in 2023 to 5,518 queries in 2024. The Inspector General reproduced the following steep declines:
Importantly, the Brennan Center report states that “the total number of U.S. person queries … for 2024 and 2025 remain unknown” (emphasis in original) due to a filter function discovered by DOJ’s National Security Division in 2024, which involved agents having not tracked some queries. That, indeed, was a problem. But, the Inspector General report explains that the issue was resolved by early 2025; and the 2025 totals of USP queries were thus largely unaffected and were close to the same low number of 2024 (7,413 in 2024).
Finally, no one is suggesting that, in an undertaking as large and complex as the Section 702 program, errors will be eradicated, but the DOJ OIG found no evidence of systemic noncompliance.
Brennan Center: It is a “myth” that a warrant requirement for backdoor searches would harm national security.
No, this is not a myth. Among Section 702’s most valuable characteristics are its speed, agility, and flexibility in providing early warnings of threats to the United States. Whether those threats be terror attacks, counterterrorism responses, narcotics interdiction or malicious cyber intrusions, speed is essential to the nation’s ability to address these threats. As the President’s Intelligence Advisory Board observed during the last reauthorization debate, a warrant mandate is “unjustified,” “impractical,” and leav[es] America “significantly less safe.” Similarly, a joint statement by senior U.S. national security officials submitted to Congress in 2023 stated, “imposing a warrant requirement … would force the government to turn a blind eye to threat information that it had lawfully acquired, with potentially grave consequences to our nation’s security.”
Brennan Center: FISA was enacted to require a warrant (or a FISA order) when the government seeks foreign intelligence.
This conflates two distinct statutory frameworks. FISA does require an order (under Title I) when Americans are the targets of surveillance conducted for foreign intelligence purposes – but Americans are not, and cannot lawfully be, the targets of Section 702 surveillance which can only be directed at foreigners reasonably believed to be outside the United States. In fact, in FISA Section 702(c)(4), Congress specifically exempts Section 702 collection from the court order requirements found in Title I of FISA. It is hard to see what applied from the very different context of targeting American agents of foreign powers for surveillance under Title I.
Brennan Center: A warrant requirement would create only a modest burden on the courts.
The basis for this assertion is unclear. Thes FISC is comprised of 11 part-time, rotating judges in the sense that all continue to serve in the districts in which they were confirmed. Previous suggested changes that would result in an increase in the FISC’s workload have met with understandable concern from the FISC. In 2025, reportedly 28% of the 7,413 USP queries produced results. That would be over 2,000 warrants for the FISC judges to handle. That said, we do not know what percentage of those results is only metadata information, for which legislative reform proposals do not require a warrant. In sum, it presumes far too much to claim the burden would be modest.
Brennan Center: Section 702 is not a “must pass” law because surveillance may continue until March 2027 under a “grandfather” clause.
False. While surveillance under existing approved certifications may continue for the duration of those certifications (generally, one year), the uncertainty of Section 702’s legal status would provoke considerable disruption – particularly with respect to the electronic communications service providers who are essential to Section 702’s operation and whose participation is essentially immunized under Section 702(i)(3). As Section 702 approached its sunset date in April 2024, two “major” communications carriers informed the government that they would stop complying with Section 702 directives if the statute expired. Section 702 is structured to operate with the assistance of U.S. communications carriers and cannot function without that assistance. What’s more, it is very easy to imagine a severe need for a new certification during the “grandfather” period but it would be disallowed, leaving the U.S. intelligence community hamstrung.
Conclusion
Given the importance of Section 702’s intelligence capabilities to the nation’s security, Congress should resist allowing the program to be fettered by a debilitating and unnecessary warrant mandate. The Brennan Center “myths” and “facts” are largely inverted, and Congress should reject this latest attempt to attach conditions that the Fourth Amendment does not require
