Two Bills and the Congressional Divide Over Section 702 Reauthorization

The clock is ticking on reauthorization of Section 702 of the Foreign Intelligence Surveillance Act. It’s easy to overlook the simmering legislative debate amid the wars, natural disasters, and political drama that have dominated news cycles since Congress returned from its August recess. Even now, heading into December, Washington is focused on looming fights over spending on Ukraine and Israel and border security, narrowly skirted a government shutdown, and may soon consider expelling one of its Members.

Outside of the intelligence community, the atmosphere around 702 seems like anything but a crisis.

If you’re feeling a certain sense of déjà vu, you’re not wrong. Our national conversations on electronic surveillance have in recent history been fraught with tension and have come down to the wire. In 2017, during the most recent Section 702 reauthorization, Congress had to buy itself additional time after missing the original deadline. Two years earlier, in 2015, Congress tempted a lapse of different surveillance authorities during its labored consideration of the USA Freedom Act. Even now, the inevitable discussion of a short-term extension of the authority is beginning to percolate as a way of buying time.

These reauthorizations have revealed a predictable rhythm: Congress looks into the abyss; figures out a solution at the last minute; and then starts a countdown clock for five years before doing it all over again. Sometimes, we remind ourselves of the risk that things might prove different this time—-that we should not be so sanguine about the present cycle. This leitmotif inevitably presents just before the eleventh hour compromise heads off anything too terrible.

But well aware that declaring “We should not be so sanguine about the present cycle” is precisely what the regular script calls for right now, we should not be so sanguine about the present cycle.

In prior eras, the countdown to 702’s lapse hasn’t coincided with the House of Representatives lacking a speaker for three weeks at a key time, the government repeatedly poised to shut down, and brinksmanship over key national security priorities with respect to foreign aid expenditures. And, adding to the political jeopardy of the moment, previous reauthorization fights haven't involved a wholesale abandonment of surveillance authorities by the MAGA right or an alignment between the MAGA right and civil libertarians in broad suspicion of FISA.

Perhaps most importantly, in the past, the government could rely on a public case about the program’s intelligence value to justify reauthorization. This time around, while the case for the program’s importance is almost universally accepted, it’s hardly a foregone conclusion that such acceptance will move Congress to reauthorize Section 702 in accordance with the government’s preferences.

In 2023, mere effectiveness and national security necessity may not be enough. 

Over the past few weeks, the Senate has unveiled two bills that attempt to take up the task of reauthorizing Section 702 of the Foreign Intelligence Surveillance Act (FISA). The proposals from—from Intelligence Committee Chairman Mark Warner and Vice Chair Marco Rubio, on the one hand, and from Sen. Ron Wyden, on the other—represent two dramatically different takes on the key question presented. That is, whether FBI compliance issues deserve precise fixes in process and culture or sweeping reform to the underlying surveillance authority. For its part, the Biden administration remains hostile to imposing a warrant requirement for FBI access to lawfully collected Section 702 information. Sen. Warner’s bill respects this position, rightly in our view. Sen. Wyden’s goes the other way. No bills from the House have emerged, although the House Intelligence Committee tipped its hand with a recent report outlining potential changes to the FISA process. With Section 702 expiring in December, the two Senate bills and the House report underscore just how far Congress still has to go in finding consensus on this issue.

Tick, tick, tick, tick, tick.

How We Got Here

One problem this time around is that while lots of different political actors have decided that 702 is a problem, they don’t share a reason. Indeed, in Washington, there are many different reasons to be dissatisfied with the current state of FISA—everything from supposed “weaponization” of the FISA process against Republicans to highly technical compliance issues at the FBI. Depending on one’s political leanings, in fact, the list of grievances could hover anywhere between as few as 1 and as many as, oh, say, 278,000.

How one feels about 702 reauthorization depends on which number one focuses upon and what  one believes is the appropriate remedy for the problem the number conveys.

Let’s start with the larger number.

This past spring, the Foreign Intelligence Surveillance Court (FISC) released an April 2022 memorandum order and opinion identifying significant compliance issues that have plagued the FBI’s handling of 702 queries since 2018. The FISC opinion revealed in part that the Bureau had conducted “in excess of 278,000 non-compliant . . . queries of raw FISA-acquired information.”

Under Section 702, the intelligence community is authorized to collect material against targeted subjects reasonably believed to be overseas under broad, basket procedures approved by the FISC but not specific to any individual target. These intercepted communications, generally produced by platform companies operating in the United States, are then stored in giant databases and can be retrieved using database queries by intelligence community components, including the FBI. While the surveillance must be conducted against targets beyond U.S. borders, it routinely includes communications with parties inside the country—including U.S. persons subject to the protection of the Fourth Amendment for the simple reason that targets abroad routinely communicate with non-targets domestically. These communications can then be queried by the FBI using identifiers associated with U.S. persons. 

The FISC memo showed that the Bureau had fallen into a “pattern of conducting broad, suspicionless queries that violate the requirement that its queries . . . be reasonably likely to retrieve foreign intelligence information or evidence of a crime.” On a large number of occasions, the Bureau had inappropriately queried its holdings for U.S. person information outside of the limited nature that Section 702 otherwise authorizes.

The scale of the issue constituted a major blemish on the bureau’s record. For example, between 2020 and 2021, the FBI ran the names of suspects from the Jan. 6 insurrection and protestors from George Floyd-related demonstrations.

That said, it’s important to be precise about what happened, what didn’t happen, and why.

The querying standard governs how the government can search information that it has already lawfully acquired under Section 702. At its most basic level, a FISA-related query must be reasonably likely to return foreign intelligence information or evidence of a crime. Part of the problem was that the FBI and the Justice Department (DOJ) had divergent interpretations of “reasonably likely.” The FBI proceeded under its less stringent interpretation. As a result, some queries were later found to lack a clearly articulated reason or a reason that sufficiently supported a Section 702-related query.

The differing interpretations spoke to the absence of a common understanding between the bureau and the department. The Justice Department’s standard, endorsed by the FISC, required a specific factual basis to believe that a query would return foreign intelligence information or evidence of a crime before a query could be conducted. At the FBI, on the other hand, agents and analysts would often run queries of U.S. person information as part of baseline checks to see what information might exist, in the first instance, within the Bureau’s holdings.

The staggering number of inappropriate queries also partially stemmed from the Bureau’s use of “batch queries” against whole groups of U.S. person identifiers. Batch queries allowed FBI personnel to more quickly identify connections among datasets of identifiers; for example, investigators might run a query that included the listings of an address book belonging to a suspected terrorist’s cell phone. But under the rules, each entry in a batch query still had to be “reasonably likely” to return foreign intelligence information or evidence of a crime. In one representative instance of a batch query that fell short, FBI personnel queried 19,000 donors to a congressional campaign that the bureau was monitoring for its foreign ties.

Other errors resulted from how the Bureau’s databases were structured technically.

As part of deconstructing the “wall” between law enforcement and intelligence—the institutional barriers that contributed to unconnected dots before September 11, 2001—FBI queries until 2021 were automatically run through multiple systems, including ones that housed unminimized Section 702 and other FISA datasets, so personnel could more easily capture all potentially pertinent information. This default-rule required FBI personnel to opt-out if they did not want to search 702 information, even though FISA-related queries are subject to a heightened standard than other FBI databases.

Another technical factor that drove compliance issues was the querying interface in Bureau systems, which allowed FBI personnel to choose from a pre-set menu of categorical justifications, rather than seeking case-specific justifications for each query.

These explanatory factors are not meant to muddy the key takeaway, which is that the FBI fell short of the compliance standards that accompany its access to sensitive information as to U.S. persons. That said, it is important to understand that the compliance issues did not reveal an agency bent on purposefully violating U.S. civil liberties. The President’s Intelligence Advisory Board (PIAB) reinforced this point in July, when it found “no evidence of willful misuse of these authorities by [the] FBI for political purposes.” Rather, the compliance issues underscored an incorrect understanding of the appropriate querying standards and systems designed to query FISA holdings based on overbroad categories.

How Congress reacts to these compliance issues is a key part of what kind of new law it considers appropriate. 

Carter Page

On the political right, there’s only one major reason to rail against FISA, and it actually has almost nothing to do with Section 702. That reason is named Carter Page—the 2016 Trump campaign adviser who got wrapped up in the Russian investigation and was subject to FISA surveillance as a result. The FBI’s failures in the Page FISA request and its reauthorizations, detailed by the Justice Department’s inspector general, generated a political firestorm. It was a moment of reckoning for the Bureau’s relationship with FISA.

As one of us reported at the time:

[A]gents failed to include in their initial surveillance application information about the target that would have undermined their showing of probable cause—including that Page had previously been an operational contact of the CIA. They overstated other things that tended to support the application, including the importance and reliability of Christopher Steele’s prior work with the FBI—which had the effect of presenting the material on Page in Steele’s reporting as more reliable than it actually was. As new information emerged that further cast doubt on Steele’s reporting, they failed to include this information in the renewal applications, leading to an ever-increasing list of errors and omissions in the Justice Department’s communications with the Foreign Intelligence Surveillance Court (FISC). In the most egregious incident, an FBI attorney is alleged to have altered an email before passing it on to the Justice Department with a change that negated Page’s prior source relationship with American intelligence.

Central to the Page fiasco was the Bureau’s failure to follow what are known as the Woods Procedures, which were designed to ensure that representations made in FISA applications are “scrupulously accurate.” A Woods File documents the basis for each factual assertion made in each FISA application; it is an essential component of the application’s credibility with the FISC. In a given Woods File, case agents will provide the background material that substantiates an assertion made in the overall FISA application. Additionally, if the application is based on reporting from a human intelligence source, the case agent will coordinate with the source’s handling agent to document the source’s reliability in the Woods File.

Or, at least, that’s how it’s supposed to work.

The public later learned that the FBI’s noncompliance with these procedures was not unique to the Carter Page application. In fact, the Bureau had committed widespread errors in its handling of Woods Files across a large sample of FISA applications. Few of these errors led to material misrepresentations in FISA applications. But it was still a very bad look.

These errors took place in the context of Title I FISA warrant applications, not Section 702 implementation, but congressional Republicans have nonetheless seized on these issues to oppose 702 reauthorization on grounds that FISA itself is being trained on conservatives. Earlier this spring, the House Judiciary Subcommittee on Crime and Federal Government Surveillance held a hearing entitled, “Fixing FISA: How a Law Designed to Protect Americans Has Been Weaponized Against Them.” Chief among the complaints of subcommittee chairman Rep. Andy Biggs (AZ-05) was how “President Obama’s FBI violated its authorities under FISA using FISA as a pretext to illegally spy on Trump campaign associates in an attempt to affect another election.”

In conflating Title I with Section 702 under the banner of weaponized government, congressional Republicans have potentially deprived reauthorization of a key conservative bloc of political supporters without whom it is not wholly clear how it passes.

Even conservatives who support the program and do not engage in this conflation are apt to be more cautious about it as a result of the FISA lapses in 2016—the theory being that adjacent violations in traditional FISA require broader reforms of FBI culture, organization, processes, including as to 702. As the Republican PCLOB members wrote, "While many of the most egregious recent violations concerning the 2016 presidential transition did not involve Section 702, Members [Beth] Williams and [Richard] DiZinno recommend that additional safeguards be enacted to ensure appropriate oversight over the program."

From the Government’s Perspective

Note that one thing nobody seems to doubt is that the 702 program is effective. Indeed, the intelligence value of the program is no less compelling than it has been during past reauthorizations. Even the government watchdogs that have critiqued aspects of the 702 program acknowledge the program is valuable.

These acknowledgments first came in July from the PIAB, which assessed that 702 information fed into intelligence streams “likely to inform every substantial national security decision our leaders make, now and in the future.”

Then, in September, the Privacy and Civil Liberties Oversight Board (PCLOB) chimed in with a split report, finding very little unanimity except in the idea that “the United States is safer with the Section 702 program than without it.”

Congress usually needs little more than that.

In this case, however, the administration is offering at least a little more than the program’s effectiveness. The administration also agrees that it needs to fix Section 702 compliance issues. Since 2021, the administration has undertaken to close some technical and training gaps that resulted in some of the glaring compliance issues, and it’s willing—eager, even—to write some of these new policies into law.

For example, and as PCLOB acknowledged, the Bureau now requires its personnel to “opt-in” for querying FISA databases (eliminating the default rule) and no longer offers pre-set categorical justifications for queries (ensuring that queries have a case-specific, factual basis). The Bureau has also instituted additional levels of approval for batch-queries.

These reforms may underscore recent results. The intelligence community reported in April that FBI’s U.S. person queries fell by over 95 percent between 2021 and 2022. The administration views these results as evidence that the Bureau is improving its culture of compliance and that a warrant requirement for data queries of U.S. persons is the wrong cure.

The Civil Libertarian Bottom Line: The Warrant Requirement

Repeat players of the reauthorization process will also recognize some familiar calls for reform from left-leaning privacy proponents. Whereas the government sees the solution to compliance problems as lying in reforms to specific procedures and codification of new rules in law, civil libertarians have renewed their demand that the government should need a court order before it can query U.S. person data from the 702 database.

Incidental collection against U.S. persons lies at the heart of this concern. Under its Section 702 authorities, the government can electronically surveil non-U.S. persons located outside the United States without a warrant. Often, as the Intelligence Community describes, these non-U.S. persons targets are either communicating with or referencing U.S. persons. All of this information gets swept up in the 702 dragnet.

Privacy proponents, including Elizabeth Goitein and Jake Laperruque, argue that Section 702 should include a warrant requirement before the government can mine its trove of incidental collection for U.S. person information. The current system, as Goitein noted in 2017, gives the government too much leeway to indirectly target U.S. person communications by collecting against the foreign party and then using queries as a backdoor for U.S. person information. Although Section 702 prohibits this practice of reverse targeting, Goitein sees the absence of a warrant requirement as a “bait and switch” around the Fourth Amendment.

In light of Section 702’s broader political jeopardy, this faction clearly has more negotiating strength than it did last time around. Its argument got a boost in September, when a majority of the PCLOB endorsed a modified warrant requirement in its report. The PCLOB sought to anticipate concerns that a blanket warrant requirement—covering all FBI U.S. person queries—would prove “unduly burdensome” on the courts. Instead, the PCLOB majority recommended,

prior to seeking FISC approval, government analysts be allowed to search the Section 702 databases without reviewing the results in order to first determine whether there is a hit on a particular U.S. person query term. Congress should allow the government to use internal procedures, under standards at least as rigorous as those in place today, to make such a determination without having to obtain FISC approval. If there is a hit on the U.S. person query term, the government would then need to request FISC approval to retrieve the content of those results. (emphasis added.)

But the PCLOB did not speak with one voice on this matter. The PCLOB’s two Republican members declined to endorse the warrant recommendation, pointing out that interposing judicial review “would most negatively impact the most important and urgent queries—the ones that show a connection between foreign targets and U.S. persons, the ones that the FBI must review as quickly as possible.”

Former FBI General Counsel Jim Baker has also explained why a warrant requirement is, in effect, totally unworkable for foreign intelligence information. The government would not be able to meet a probable cause standard required for criminal warrants. Agencies would lose speed and agility in responding to aggressive and adaptive foreign adversaries, especially during cyber incidents. Finally, the volume of queries would overwhelm the system. Granted, Baker’s assessment predated the modified warrant requirement recommended by the PCLOB majority, which would call for the FISC to evaluate applications under the current “reasonably likely” standard and not a probable cause standard. His concerns about speed and agility, though, still resonate, especially in cases where the government is running defensive queries to identify or warn potential victims of a cyber intrusion.

In short, the current FISA 702 debate currently divides Congress into four distinct camps:

1. Civil libertarians whose core requirement is a warrant requirement for access to 702 data;
2. The intelligence community’s backers, who are willing to address compliance issues using a variety of increased accountability and regulatory measures but who draw the line at judicial permission for government access to legally collected material;  
3. The coalition of hard-core civil libertarians and burn-it-all down conservatives who oppose 702 altogether; and
4. A moderate middle who will agree to anything the three camps above can settle on. This group used to be the vast majority of both houses but is shrinking notably. 

There is no way to square the circle. Someone has to win here and someone has to lose.

Senators Warner and Rubio, on the one hand, and Wyden, on the other, have picked very different winners and losers.

Senator Warner: Target the Compliance Issues

Sen. Warner’s bill has two main components. The first half would reauthorize Section 702 through 2035—which would offer the government a good long time before it would have to go through the agony of reauthorization again. This portion largely addresses how the FBI interacts with U.S. person information under Section 702 and, notably, does not contain a warrant requirement. The second half of the bill offers broader reforms of Title I of FISA.

Instead of proposing a warrant requirement, the bill simply eliminates the FBI’s authority to query its 702 holdings solely for purposes of discovering “evidence of a crime.” Since 2018, the FBI has been required to obtain a court order before accessing the results of 702-related queries designed solely to find evidence of a crime. The provision in Warner’s bill would do away with this authority completely. In doing so, Warner’s bill would not affect queries designed at least in part to surface foreign intelligence information. As these represent nearly all queries under Section 702, the “evidence of a crime” ban is an almost-entirely optical exercise—banning a category of queries that is all-but an empty-set anyway. Under Warner’s bill, queries with a foreign intelligence collection component, alongside its criminal investigative component, would still proceed under current rules.

The bill would also codify several policies undertaken in recent years to address FBI compliance issues. The attorney general and director of national intelligence would be required to institute mandatory training standards for FBI personnel, for example; and it would codify the “opt-in” model for Section 702 queries within FBI systems, attorney approval of FBI batch jobs, and the requirement of a written statement of the specific factual basis for each query. An FBI senior leader would also have to approve queries related to political figures, religious officials, or journalists.

The bill also contemplates stronger congressional oversight of FBI querying. The FBI would submit annual reports to Congress on several categories of Section 702 queries, including the number of U.S. person queries, batch job queries, and evidence-of-a-crime queries. Additionally, with limited exceptions, the FBI would be required to notify Congress about queries likely to identify any of its members. Senator Warner’s bill also includes an exception for consensual queries and new requirements for queries that vet non-U.S. persons for U.S. travel.

The Warner-Rubio bill also reforms some aspects of the traditional FISA application and adjudication processes—-clearly a response to the Carter Page issues.

The bill strengthens the accuracy requirements of FISA orders, for example. FISA court applications would require a certification from the applicant that the Justice Department is apprised of any information that would raise doubts on the accuracy of assessments contained in the application. The use of information from political organizations in applications for Title I and Title III FISA orders would be subject to more stringent requirements. The bill also includes stronger penalties for false declarations, unauthorized use of FISA information, and unauthorized disclosure of a FISA application.

Finally, the bill provides the FISC with more independent legal advice and expands the range of circumstances in which amici presumptively participate in FISC proceedings. The FISC would be required, for example, to appoint amici whenever it reviewed applications that present a “novel” interpretation of law or propose targeting a U.S. person during a sensitive investigative matter unless the court found it was not appropriate and, in sensitive cases, some emergency required otherwise. Notably, this is a provision the administration has opposed in the past. Its inclusion in even Warner’s bill this time around shows how far the debate has moved.

Representative Turner: Extensive Compliance Reforms to Prevent Another Carter Page

On November 16, the House Permanent Select Committee on Intelligence (HPSCI) released a report foreshadowing the House Intelligence committee’s own likely legislative proposals on FISA and Section 702 reforms. The committee’s Majority FISA Working Group, comprising Chairman Mike Turner and Republican Representatives Darin LaHood and Brian Fitzpatrick, authored the report, whose recommendations focus on Section 702, traditional FISA applications, and FISC operations. The report also suggests statutory updates that would strengthen electronic surveillance of drug cartels, fentanyl traffickers, and “lone wolf” terrorists.

The FBI’s recent Section 702-related compliance issues drive the House majority’s proposed amendments. As with the Warner-Rubio bill, there is no call for a general warrant requirement for queries.

Rather, the report includes heightened penalties for compliance violations or abuse and creates criminal liability for intentional leaks of information obtained under Section 702. Additionally, the report suggests strengthening Congress’s oversight of Section 702 surveillance, including notification requirements of Section 702 querying of members of Congress and reporting requirements detailing U.S. person query metrics.

The House Intelligence Committee majority, while not imposing a warrant requirement, would significantly tighten FBI’s querying procedures. Several of the report’s reforms codify current policy, including attorney approval for batch job queries and heightened training requirements. But, if enacted, many FBI personnel—“more than 90%”—would lose their ability to authorize U.S. person queries, as these would be limited to a subset of supervisors and attorneys. The FBI would have to obtain a probable cause warrant for conducting evidence-of-a-crime queries, and queries designed to “suppress” First Amendment activities would be expressly prohibited.

The working group constructed its Title I reforms with an eye toward preventing another Page situation. The report includes heightened penalties for intentional misconduct, false declarations, and intentional disclosures. Additionally, several amendments are designed to improve accuracy and completeness of FISA applications with additional requirements for the underlying factual statements and certifications. Use of information from political organizations or press reports in FISA applications would be subject to restrictions and require corroboration.

The report’s FISC reforms would also seek to improve the court’s transparency and require appointment of amicus curiae in the annual Section 702 reauthorization process. Congress would receive more tools to carry out oversight of FISC operations, including by receiving copies of FISC transcripts and permitting certain congressional officials to attend FISC proceedings.

Senator Wyden: Sweeping Change Begins with a Warrant

If Sen. Warner’s bill represents the opening position of the intelligence community camp, Sen. Wyden’s bill represents the civil libertarian camp’s hope, incorporating Section 702 reauthorization into an extensive package of reforms addressing FISA, the Electronic Communications Privacy Act, and other intelligence oversight topics. The bill would fundamentally change how the government collects, queries, and uses U.S. person information under FISA.

Sen. Wyden’s bill takes direct aim at how the government may query and use U.S. person information derived from Section 702 collection. Principally, the bill would prohibit U.S. person queries under Section 702 in the absence of a probable cause warrant. The bill carves out exceptions during emergency situations. This provision would institute judicial approval over nearly all U.S. person queries, through a Title I or Title III FISA order or, as a result of the bill, a criminal warrant. Section 702-related queries for communications metadata would have a lower threshold of querying requirements.

The bill would also restrict the government to using 702-acquired U.S. person information during proceedings or investigations related to a specific type of national security threat and would institute a five-year retention limit on U.S. person information collected under Section 702.

The bill would also change several aspects of Section 702 collection. First, it would require that obtaining foreign intelligence information is the sole purpose behind a Section 702 acquisition. This provision would thus strengthen current law, which only requires that obtaining foreign intelligence information be a significant purpose. The bill would also prohibit any resumption of “about” collection—the search of 702 data not for material reflecting communications involving a U.S. person but for communications about that person—and strengthen the statutory prohibition against reverse targeting of U.S. persons using Section 702 collection. Finally, the government would be required, prior to compelling technical assistance from an electronic communication service provider, to demonstrate to the FISA court that such assistance is necessary, narrowly tailored, and does not pose an undue burden to the provider.

More broadly, Sen. Wyden’s bill replaces sections 703, 704 and 705 of FISA with a blanket prohibition against intentional targeting of a U.S. person, regardless of location, or a person reasonably believed to be located in the United States to acquire foreign intelligence information when the person has a reasonable expectation of privacy or a warrant would be required. The bill allows such collection with a warrant or in an emergency. Sen. Wyden’s provision resembles the current structure of FISA section 704, which, according to the Office of the Director of National Intelligence, requires an individualized court order “when the government seeks to target a U.S. person overseas under circumstances in which the U.S. person has a reasonable expectation of privacy and a warrant would be required if the acquisition were conducted in the United States.”

Sen. Wyden’s bill also includes numerous reforms of the FISA application and adjudication process. All FISA applications would require a description and certification of accuracy procedures used by the applicant. Amici would be granted access to classified information and permitted to participate in a wider array of FISC proceedings. Novel or significant FISC decisions would be subject to declassification within 180 days and the FISC itself would receive jurisdiction to hear claims for access to records. 

In addition to these procedural elements, the bill would sunset collection authority under Section 215 of the PATRIOT Act and allow parties to bring suit for a surveillance-related injury-in-fact, abrogating the state secrets privilege.

Where To Go From Here

Reauthorizing 702 is not really optional. Or, rather, it’s optional only in the sense that raising the debt limit is optional. Congress always has the option of catastrophically self-destructive inaction, but in the department of plausibly-responsible government, there is no universe in which something like the authority that exists today does not remain in force after the turn of the year.

The only serious policy question on the table is whether or not the current regime should contain some kind of warrant requirement.

To be blunt on this point, the advocates of the warrant requirement have not made a persuasive case either for its necessity or its utility. Leaving aside the conceptual incoherence of requiring judicial review for the executive branch to look at information already lawfully in its possession as a result of court-supervised surveillance activity, a warrant requirement is also non-responsive to the compliance issues that have arisen—responding mostly to imagined civil liberties issues that have not arisen. The idea of a warrant is most responsive to the concerns about so-called “back door searches” or reverse-targeting of U.S. persons; but the compliance problems that have actually arisen do not reflect efforts to game the system and get around rules restricting access to U.S. person data; they involve errors and misunderstandings of rules. And they are correctable by far-less cumbersome means than the imposition of judicial review over queries.

One consequence of all of the recent disputes over FISA is that the number of FISA applications has plummeted over recent years. Here, we cite the ODNI Annual Statistical Transparency Reports regarding IC FISA usage:

Section 702

• The FBI’s U.S. person queries conducted under Section 702 fell by over 95 percent between 2021 and 2022.
• Other IC component (CIA, NSA, NCTC) U.S. person queries conducted under 702 decreased from 8,406 in 2020 to 4,684 in 2022.

“Probable Cause FISA” (requiring a Title I, Title III, Section 703, Section 704 court order)

• The IC’s total number of FISA Orders has decreased from 524 in 2020 to 337 last year. These figures continue a longer-term trend. In 2017, when Section 702 and surveillance authorities were previously under intense public scrutiny, the IC had 1,437 of these orders on record. (And even the 2017 figure represented a five-year low from 2013.)
• U.S. persons only comprised roughly 12% of FISA “probable cause” targets in 2022, down from about 23 percent in 2020.

Civil libertarians might find this trend salutary, less surveillance being all for the good. We are less categorical. A decline in FISA’s use can mean a decline in counterintelligence against Russian and Chinese targets. It can mean less visibility into terrorist groups. It can mean less ability to counter foreign malign influences domestically. It can mean less effective cybersecurity. These downstream effects may underscore a growing perception of professional risk that IC personnel associate with using FISA, even when it may be perfectly appropriate to do so.

From this vantage point, the decreases in querying and surveillance orders give us pause because the pool of foreign targets is actually growing. Since 2020, the estimated non-U.S. person targets under Section 702 has consecutively increased. The number of FISA targets subject to court orders also increased from 2021 to 2022. Additional non-U.S. persons are driving this trend, given the decline in U.S. person “probable cause” targets.

Even the Warner-Rubio bill would further gum up the FISA works with new procedures and bureaucratic mechanisms that would make the process less agile.

Still, something like Warner-Rubio is probably the best the administration can hope for but the accretion of these rules by way of holding off a revolution in surveillance law like the proposed warrant requirement itself takes a toll over time.

Tick, tick, tick, tick.