Published by The Lawfare Institute
in Cooperation With
More than 200,000 foreign targets a year. “Irreplaceable” insight. The National Security Agency’s “most significant tool” for disrupting terrorist threats. All achieved at low cost and with virtually no operational risk.
The law that makes this possible, Section 702 of the Foreign Intelligence Surveillance Act (FISA), allows U.S. intelligence agencies to collect data on overseas foreign targets from communications infrastructure in the United States. Because the U.S. remains the leader in global internet services, much of the world’s data flows here, making this type of collection a bounty for U.S. intelligence (and thus for U.S. allies as well).
Unless Congress acts, however, Section 702 will expire on Dec. 31, 2023.
What happens then?
There is no comparable substitute. “Traditional” FISA, which requires lengthy applications to the secret FISA court, was used in fewer than 500 cases last year. Section 702, which doesn’t require individual court orders, can cover orders of magnitude more targets: more than 230,000 in 2021. Even a small fraction of that volume would overwhelm the FISA court.
Nor can overseas collection under Executive Order 12333 make up the difference. To be sure, Executive Order 12333 collection on these targets would be lawful; all of Section 702’s targets are non-Americans overseas. The intelligence community’s heavy use of the more constrained Section 702 suggests, however, that collecting data on home turf offers considerable advantages over collection abroad.
The Hardest Reauthorization Yet
For a decade after 9/11, surveillance laws drew reliable, bipartisan majorities in Congress. Even as memories of the attacks faded, congressional leaders mustered enough votes to renew Section 702 in 2012 and 2018. In 2018, reauthorization passed with 256 votes in the House and 65 in the Senate, though not without some scrambling.
Will the votes be there this time?
Republicans will control the House of Representatives, with a slim and fractious majority.
The Biden administration will need Republican votes. More subtly, it will need House Republican leadership to move the bill and whip support for it—something the new speaker will be reluctant to do if a majority of Republican members oppose it.
And oppose it they may. Jim Jordan, the incoming chairman of the House Judiciary Committee, said last month: “We need to make changes to the FISA process. I think we should not even reauthorize FISA, which is going to come up in the next Congress.” Jordan is a longtime skeptic of FISA and the FBI. But he’s far from alone: Many Republicans have become deeply skeptical of the national security state.
This skepticism has various causes. Some relate to FISA: most notably, the blunders in the Crossfire Hurricane investigation and the errors, misstatements, and omissions in FISA court applications to surveil Carter Page, then an adviser to Donald Trump’s presidential campaign.
Other causes, such as the Disinformation Governance Board and the FBI’s search of Trump’s Mar-a-Lago residence, do not relate to FISA but have fed a broader mistrust of national security institutions.
After the Mar-a-Lago search, Kevin McCarthy, the presumptive next speaker of the House, issued a statement saying: “I’ve seen enough. The Department of Justice has reached an intolerable state of weaponized politicization. When Republicans take back the House, we will conduct immediate oversight of this department, follow the facts, and leave no stone unturned.”
Another leading Republican, Senate Judiciary Committee Ranking Member Chuck Grassley of Iowa, has lambasted the FBI for partisan bias on the part of senior supervisory agents. In one of many oversight letters on the topic, he said: “I remain very concerned that political bias by a select group of Justice Department and FBI officials has infected the Justice Department’s and FBI’s usual process and procedure to open and pursue high-profile and politically charged investigations.”
This skepticism isn’t simply an elite phenomenon: Republican voters have grown progressively more mistrustful of national security institutions, especially the FBI. That mistrust now colors views of surveillance laws, making the reauthorization of Section 702 much more challenging.
The Path to Reauthorization
Even with an all-out push, there is no guarantee that reauthorization will pass. But there are several steps that intelligence community leaders and other executive branch officials can take now to make reauthorization more likely.
1. Designate leaders for the reauthorization process.
Past FISA reauthorization efforts have struggled because of a lack of clarity about who within the executive branch leads the reauthorization process. If it has not already done so, the administration should clearly identify quarterbacks to lead congressional outreach and drive the declassification of necessary materials.
2. Publicly acknowledge Republicans’ concerns and express openness to improving safeguards around FISA, including Title I.
The basic problem is a lack of trust. Republicans believe that national security institutions have misused their power to intervene in political campaigns and public discourse. They do not hear their concerns acknowledged by those who wield these authorities. Instead, errors are downplayed or “contextualized” and objections dismissed.
The administration, Democrats in Congress, and nonpartisan intelligence experts should set aside for a moment whether they believe Republican critiques of the national security state are correct. Their conviction that these critiques are wrong has led them to mistakenly assume that these critiques are also irrelevant and can thus be safely ignored.
Raw politics now dictate that they cannot be ignored. If reauthorization depends on Republican votes and a Republican speaker’s engagement, Republican concerns matter. Increasing numbers of Republican voters see national security institutions as hostile to their beliefs. Their representatives will not vote to empower these institutions without some assurance that these concerns are being addressed.
Simply acknowledging that these concerns exist, and being willing to discuss them, is the first step.
3. Do not dismiss Carter Page.
When talking with non-Republican intelligence experts, I often hear a version of the following argument: Republicans’ concerns about former Trump campaign adviser Carter Page have nothing to do with Section 702, because Page was surveilled under a different part of FISA (Title I).
That is true but unsatisfying: If safeguards at the FBI, the Justice Department, and the FISA court failed to protect Page, then why, skeptics would naturally ask, would they prevent misuse of Section 702? This mistrust centers not on legal texts, but on the institutions and people that use these authorities.
Dismissing concerns about the Page case and allegations of partisanship within the FBI also signals that past misconduct is not being taken seriously but, rather, explained away. A better approach would be to acknowledge the mistakes in the Page case and express openness to discussing additional safeguards.
4. Quietly engage with key Republicans on the Hill.
For all those who will likely oppose Section 702, there remain many powerful Republicans who support it and will work to keep it alive. For example, Republicans on the House Intelligence Committee, led by incoming Chairman Mike Turner (R-Ohio), have already begun preparing for reauthorization. Turner has assigned a working group of committee Republicans, led by Rep. Darin LaHood (R-Ill.), to conduct a “comprehensive review” of issues related to the process.
Conservative, pro-intelligence Republicans like Turner, Rep. Michael McCaul (R-Texas), and LaHood on the House side, and Sens. Marco Rubio (R-Fla.) and John Cornyn (R-Texas) on the Senate side, will be best positioned to gauge what measures might persuade reluctant but “gettable” Republican votes. The path to reauthorization runs through them.
5. Find new ways to illustrate Section 702’s value.
In past reauthorizations, intelligence community officials counted on rallying members with classified briefings shortly before the vote. Typically, anecdotes about counterterrorism successes were seen as most effective.
Changed circumstances, however, call for new tactics. Fear of terrorism is not the political catalyst it once was. The intelligence community should not rely principally on counterterrorism vignettes or classified briefings to drive votes for reauthorization, for a few reasons.
First, the political stakes for Republicans have risen. Where the Republican base is skeptical, it will be risky for representatives to buck their voters on the basis of classified evidence that those voters cannot see.
Second, global terrorism has become less salient as a political factor. In March 2020, Congress allowed three terrorism-related provisions of FISA to lapse. They have not been revived. Those provisions were not nearly as important as Section 702. But the first sunset of any counterterrorism power since 9/11 should nonetheless be a warning that fears of terrorism do not drive votes as they once did. At least among Republicans, China, not terrorism, looms largest.
What should the intelligence community do instead?
Most importantly: to the extent possible, the intelligence community should declassify information about how Section 702 helps combat threats other than terrorism that are more politically salient. Classified briefings can describe these non-counterterrorism successes to members in greater depth.
6. Prepare to negotiate.
In past reauthorization cycles, agencies have been loath to open the door to statutory changes—until it becomes obvious that a bill won’t pass without them, at which point the haggling begins. Typically, the government’s opening position has been “clean” reauthorization without any substantive changes.
There is little chance of a clean reauthorization this time (though temporary extensions are always a possibility). In theory, opening with no concessions leaves room to bargain. But it also risks bolstering the perception that officials do not take Republican concerns seriously.
Republicans are split. Some see the value of Section 702 and will push hard for it. Some are beyond persuasion. Others could go either way—and their votes may be decisive. How can those potentially “gettable” Republicans be persuaded?
As Stewart Baker has argued, the best way to make this vote politically viable for Republican fence-sitters is to present it as an opportunity to remedy political abuses of surveillance powers. Letting Section 702 lapse would benefit foreign adversaries; passing a bill is a chance to protect Americans’ rights.
To that end, supporters of reauthorization should begin preparing options for statutory changes and other moves that can assuage concerns and win votes.
Politically constructive tweaks might include the following:
- Adding to FISA’s definitions a category encompassing highly sensitive investigations of U.S. persons, modeled on the FBI’s internal “Sensitive Investigative Matter” designation. Under the Attorney General’s Guidelines for Domestic FBI Operations, Sensitive Investigative Matters include matters involving domestic public officials or candidates, religious organizations or leaders, and journalists, along with other investigations carrying heightened sensitivity. Creating a statutory category would enable Congress to apply heightened safeguards to the most sensitive FISA applications involving Americans.
- Requiring the appointment of an amicus curiae in cases presenting heightened civil liberties concerns. The creation of the FISA court’s panel of amici curiae, a recommendation from the Privacy and Civil Liberties Oversight Board’s 2014 report on Section 702, has worked well. As currently drafted, however, the statute requires that the court consider appointing an amicus only when a matter “presents a novel or significant interpretation of the law, unless the court issues a finding that such appointment is not appropriate.” Other matters in which an amicus appointment could be made mandatory include the following:
- All Section 702 reauthorizations. As I have argued elsewhere, “[g]uaranteeing that an amicus will be appointed in this narrow, but very important, category of cases would strengthen the public credibility of Section 702’s programmatic judicial oversight.”
- Sensitive Investigative Matters (or a new statutory analogue) involving U.S. persons. These cases are not numerous, so this could conceivably be handled within the existing amicus structure.
- Some observers have argued that an amicus should be appointed for each probable-cause-based FISA application involving a U.S. person. There were 67 of these in 2021. Even at this historically low level, this many appointments would likely overwhelm the existing amicus panel, which relies on a small number of outside experts who have other full-time employment.
- Congress could also consider creating a full-time amicus panel to raise factual and legal arguments in applications involving Americans. Alternatively, the FISA court could designate certain legal advisers to perform an internal red-team function in those cases.
- Requiring the executive branch to grant American citizens equal access to any executive branch remedies available to noncitizens, such as the system created by the new Executive Order 14086. The executive order, which President Biden issued in October, offers residents of certain foreign countries a redress process for alleged violations of rules related to signals intelligence. Congress should not mandate that such a process exist, but it could require that if the executive branch chooses to create such a process for noncitizens, it must offer it to Americans as well.
- Requiring that the Justice Department conduct full-file “completeness” reviews to check for omissions in sensitive FISA applications and report to Congress on the filings. The Justice Department began conducting these reviews, which check for information in FBI files that may have been omitted from FISA applications, in the wake of the Carter Page blowup. Searching for omissions is very labor intensive, but Congress could mandate these checks in a narrow subset of FISA filings. For example, it could require a full-file review in all U.S.-person Sensitive Investigative Matter filings and in a randomized sample of other applications.
- Requiring renewal-specific findings before approving applications to renew surveillance on Americans. The flawed surveillance of Carter Page was renewed three times, suggesting that this is an area for potential improvement. Counterintuitively, most renewal applications presented to the FISA court use the same structure as initial filings. Those that I have seen did not specifically reassess their original decision in light of the information obtained through the surveillance. Requiring the court to specifically consider whether, for example, the initial surveillance alters the basis for probable cause could add rigor to the renewal process.
- Requiring renewals to be reviewed by the same FISA court judge who considered the initial application, as proposed by Rep. Chris Stewart (R-Utah) in late 2019. Logically, a judge who is already familiar with the facts will be better able to assess whether more surveillance is warranted. Potential scheduling concerns should be surmountable.
- Providing for transparency around unmasking during presidential transition periods. Respected U.S. Attorney John Bash reviewed unmasking requests during the 2016 election and the Obama to Trump transition and “found no evidence that unmasking requests were made for political purposes or other inappropriate reasons.” Bash also found, however, that “the unmasking of USPII [U.S.-person identity information] that relates to associates of a presidential campaign or to transition officials creates a risk of abuse, such as through the leak of politically damaging information to the media.” For that reason, Bash proposed that the intelligence community provide greater transparency around unmasking of campaign-related identities during presidential transitions.
Historically, unmasking has been regulated by agency procedures, not by statute, so it would be anomalous for Congress to directly impose such requirements—especially where the data was collected under Executive Order 12333 rather than FISA. Were the intelligence community to implement such changes now, however, it could generate some goodwill among Republicans in advance of the 2024 election. Congress could also require the intelligence community to provide a report on this topic after each presidential transition.
7. Consider going big.
Many features of FISA are due for rethinking. It is not technology agnostic. It relies on geographic distinctions that have been undermined by the global internet. Most glaringly, FISA makes it only modestly more difficult to surveil Americans and lawful permanent residents than temporary foreign visitors—even though these groups are quite differently situated in terms of constitutional rights, allegiance to the United States, and other interests.
Sharper distinctions around nationality would also align with broader trends in American national security. Great power competition, especially with China, now outranks counterterrorism. In that sphere, temporary foreign visitors, rather than Americans who might be radicalized and controlled from abroad, may form the population of greatest concern.
A deeper reform might include the following elements:
- Replacing the complex definition of “electronic surveillance” with a simpler approach based on familiar principles of constitutional law. Another provision of FISA, which governs overseas surveillance of overseas Americans, offers a ready model. Section 704 requires the government to obtain a judicial order based on probable cause “under circumstances in which the … person has a reasonable expectation of privacy and a warrant would be required if the acquisition were conducted inside the United States for law enforcement purposes.”
- Adding stronger protections for U.S. persons, such as those proposed in the previous section.
- Making it less burdensome to initiate surveillance of foreign governments, entities “directed and controlled by a foreign government,” and agents of a foreign power who are not U.S. persons.
- For example, Congress could consider the following:
- Creating a 702-like annual certification regime for “traditional” FISA surveillance of foreign governments, entities “openly acknowledged by a foreign government … to be directed and controlled by such foreign government,” and these targets’ officers and employees. Energy spent on these clear cases could be better used elsewhere: for example, enforcing minimization rules that protect the privacy of U.S. persons ensnared in FISA surveillance.
- Extending the maximum duration of surveillance targeting foreign powers and agents of a foreign power who are not U.S. persons.
- Making oversight of Section 702 targeting more efficient by using automation and statistically valid sampling.
- For example, Congress could consider the following:
Starker distinctions around nationality would benefit both national security and civil liberties. They would free up resources that could better be used to safeguard the rights of Americans. And they would help intelligence agencies better track people with no allegiance and only transitory connections to the United States who may be doing the nation harm.
Indeed, in many cases, stopping malicious foreign agents is essential to protecting Americans’ constitutional rights. For example, agents of the Chinese Communist Party have tried to silence the regime’s opponents here in the U.S. Effective surveillance of genuine foreign agents helps protect Americans’ freedom.
To be sure, reopening core elements of FISA would broaden the scope of potential disagreement. It might, however, create new possibilities for bipartisan deals to facilitate the reauthorization of Section 702.
Of course, this could all change between now and next December. A mass casualty terrorist attack, a Chinese invasion of Taiwan, direct U.S. involvement in Ukraine, or an indictment of former President Trump could scramble the politics of national security yet again. In the end, world events, not legislative tactics, may decide whether Section 702 survives.