Assessing the Field of Section 702 Reform (and Anti-Reform) Bills

As 2023 comes to a close, the debate over reauthorizing Section 702 remains heated. However, because of Congress’s disappointing and unnecessary “temporary” extension of Section 702 as part of the National Defense Reauthorization Act (NDAA), there is likely to be no further movement on the reform debate until early 2024—that is, if the administration does not get a new, year-long certification further punting reform debates into 2025. Ahead of further debate, let’s look at the current proposals for reauthorizing Section 702.

The Stakes and EPIC’s Criteria for Reform

Section 702 authorizes surveillance programs targeting non-U.S. persons reasonably believed to be outside the United States to acquire “foreign intelligence information” with the compelled cooperation of U.S. electronic communication service providers. Section 702 surveillance is conducted without adhering to the traditional rules of FISA and without individualized court orders by the Foreign Intelligence Surveillance Court (FISC). Under Section 702, the government collects hundreds of millions of communications a year, including an unknown—but significant—number of communications to or from Americans communicating with any of the 246,073 foreign targets.

It is Section 702’s impact on Americans in particular that has motivated conversations around reform this year. Because Section 702—by design—sweeps in Americans’ communications, agencies have “minimization procedures” for this information. However, under these rules, agencies can and do search for particular Americans’ information in Section 702 databases, turning Section 702 into a potent tool for domestic surveillance. Since the last reauthorization cycle alone, government audits and FISC opinions have revealed hundreds of thousands of improper searches, including tens of thousands of improper searches relating to civil unrest and significant civil rights abuses.

Because of these abuses, EPIC believes there should be no reauthorization of Section 702 absent significant reform. In March, EPIC laid out its key priorities for reform as part of any reauthorization of Section 702.[1] And in June, EPIC joined a bipartisan coalition of civil society groups in endorsing a broader set of reforms consistent with EPIC’s prior key priorities. Together, these form the basis for EPIC’s evaluation of any legislation proposing to reauthorize Section 702.

We grouped our key priorities into three substantive categories: (1) reforming surveillance pursuant to Section 702; (2) reforming surveillance pursuant to Executive Order 12333; and (3) closing the data broker loophole. We can take each of those in turn.

Reform Category	EPIC Reform Priority
Reforms to Section 702 and FISA	End warrantless backdoor searches. All agencies should be required to obtain a warrant based on probable cause before conducting a U.S. person query.
	Codify stronger purpose limitations. Stronger purpose limitations should be codified for the use of Section 702 surveillance, for example: (a) Limiting targeting to foreign powers or agents of foreign powers; or (b) Codifying Executive Order 14086’s list of legitimate purposes and prohibited purposes.
	Prohibit “abouts” collection. Any reform bill should also codify an end to “abouts” collection, as well as clarify that such a prohibition applies to all programs conducted under Section 702.
	Require robust data minimization. Stricter data minimization rules must be imposed across all agencies, including but not limited to: shorter default retention periods and narrower exceptions to these default periods; stronger oversight of purging compliance; and a prohibition on the use of attorney-client privileged communications for any purpose, except to ensure compliance with purging requirements.
	Enhance oversight, accountability, and transparency. EPIC recommends any reform bill include FISA Court reforms consistent with the text of the Leahy-Lee Amendment.
	Ensuring judicial review. Any bill should prohibit “parallel construction” and require that the government notify criminal defendants any time they intend to introduce evidence that would not have been discoverable but for the use of Section 702. Any bill should also expressly prohibit include more robust requirements for notifying defendants of the government’s use of Section 702-derived evidence; and (3) it should ensure that surveillance programs are accountable in U.S. courts, including by reforming the doctrines of standing and the state secrets privilege.
EO 12333 Surveillance Reform	Parallel EO 12333 Surveillance Reforms. Surveillance reforms should extend to the activities conducted under EO 12333, including a prohibition on bulk collection, a warrant requirement for all agency queries of U.S. person information, and alignment of Section 702 and EO 12333 oversight frameworks.
Data Broker Loophole	Close the Data Broker Loophole. Congress must also prevent the government from circumventing Americans’ Fourth Amendment protections by purchasing data instead of obtaining a court order.

It should also go without saying that implicit in EPIC’s criteria is that a surveillance reform bill should not expand surveillance authorities. Therefore, we’re going to count that as its own category when assessing the bills below.

Assessing the Current Surveillance Reform Bills

Heading into 2024, there are now four different bills in Congress that would reauthorize Section 702.[2] Below is a table comparing the key surveillance reform bills against EPIC’s key priorities:

Below, we’ll summarize some of the key reforms (or lack thereof) in each, as well as any problematic provisions.

The (Almost) North Star: The Government Surveillance Reform Act (GSRA) [S. 3234/H.R.6262] [Endorsed by EPIC]

• Senate Cosponsors (as of December 21): Sens. Wyden (D-OR), Lee (R-UT), Hirono (D-HI), Daines (R-MT), Heinrich (D-NM), Lummis (R-WY), Tester (D-MT), Warren (D-MA), Markey (D-MA), Baldwin (D-WI), Merkley (D-OR)
• House Cosponsors (as of December 21): Reps. Davidson (R-OH), Lofgren (D-CA), Biggs (R-AZ), Jayapal (D-WA), Massie (R-KY), Jacobs (D-CA), Mace (R-SC), Correa (D-CA), Carey (R-OH), Chu (D-CA), Armstrong (R-ND), DelBene (D-WA), Lieu (D-CA), Hoyle (D-OR), Lee (D-CA), Doggett (D-TX), Deluzio (D-PA), Cohen (D-TN), Garcia (D-IL), Casar (D-TX)

As we noted in a previous blog post, the GSRA is a “sweeping, bipartisan, bicameral bill that would rein in the growing ecosystem of warrantless government surveillance authorities” while also reauthorizing Section 702. The GSRA’s key reforms include:

• Prohibiting, subject to certain exceptions, warrantless searches of Americans’ communications collected without a warrant;
• Bolstering the FISA Court safeguards in line with the Lee-Leahy amendment;
• Requiring that all agencies with access to Section 702 information have robust accountability measures in place to punish misuse of surveillance authorities; and 
• Prohibiting intelligence and law enforcement agencies from exploiting the Data Broker Loophole to warrantlessly buy Fourth Amendment-protected information about people in the United States en masse, including our location and internet records.

While the GSRA does not address all of EPIC’s major reform priorities, it does contain other important surveillance reforms. As we note in our quick guide to the GSRA, the bill has other vital reforms to intelligence and law enforcement surveillance, including ending a grandfather clause permitting continued surveillance conducted under Section 215 of the PATRIOT ACT, enacting safeguards for the use of cell-site simulators, and closing loopholes enabling warrantless collection of car data.

The Good: The Protect Liberty and End Warrantless Surveillance Act [H.R.6570][Endorsed by EPIC]

• House Cosponsors (as of December 21): Reps. Biggs (R-AZ), Nadler (D-NY), Jordan (R-OH), Lofgren (D-CA), Davidson (R-OH), Jayapal (D-WA), Fry (R-SC), Jacobs (D-CA), McClintock (R-CA), Chu (D-CA), Crane (R-AZ), Lieu (D-CA), Hoyle (D-OR), Tiffany (R-WI), Schakowsky (D-IL), Gooden (R-TX), Lee (D-CA), , Doggett (D-TX), Roy (R-TX), Duncan (R-SC), Collins (R-GA), Tlaib (D-MI), Mace (R-SC), Casar (D-TX), Burlison (R-MO), Good (R-VA), Bishop (R-NC), Van Drew (R-NJ), Weber (R-TX), Hageman (R-WY), Blumenauer (D-OR), Mooney (R-WV), Porter (D-CA), Del. Norton (D-DC)

EPIC has also endorsed the Protect Liberty Act, which passed out of committee by an overwhelmingly bipartisan vote (35-2) on December 6. The Protect Liberty’s key reforms include:

• Requiring a warrant for searches of Americans’ communications collected under Section 702;
• Bolstering the FISA Court safeguards and accountability measures for misuse of surveillance authorities;
• Codifying the end to “abouts” collection pursuant to Section 702; and
• Enacting the Fourth Amendment Is Not for Sale Act, which prohibits intelligence and law enforcement agencies from exploiting the Data Broker Loophole to purchase location and other sensitive information.

Although narrower in scope than the GSRA, the Protect Liberty Act’s core provisions mirror those of the GSRA. For example, the Protect Liberty Act’s warrant requirement is very similar to that in the GSRA and includes the same reasonable exceptions for consent-based searches, certain defensive cybersecurity searches, or where there is an imminent threat of death or serious bodily harm.

The Protect Liberty Act would also enact the Fourth Amendment Is Not for Sale Act, which—though not identical in form to the provisions of the GSRA—would go a long way towards closing the data broker loophole for sensitive information like location data.

The Protect Liberty Act also contains FISA court reforms that, while not going as far as those in the GSRA, are nearly identical to the Lee-Leahy amendment, which passed the Senate in 2020 by a 77-19 vote.

Overall, as EPIC’s Jeramie Scott noted upon its introduction, the Protect Liberty Act, unlike the HPSCI and SSCI bills, “is what responsible, meaningful reform looks like.”

The Bad and the Ugly: The HPSCI and SSCI Bills [S.3351/H.R.6611]

• Senate Cosponsors (as of December 21): Sens. Warner (D-VA), Rubio (R-FL), Graham (R-SC), Wicker (R-MS), Collins (R-ME), King (I-ME), Bennet (D-CO), Casey (D-PA), Gillibrand (D-NY), Rounds (R-SD), Kelly (D-AZ), Klobuchar (D-MN), Moran (R-KS), Lankford (R-OK), Whitehouse (D-RI), Manchin (D-WV)
• House Cosponsors (as of December 21): Reps. Turner (R-OH), Himes (D-CT)

While the GSRA and Protect Liberty Act comprise the reform movement in Congress, the Senate Select Committee on Intelligence (SSCI) and House Permanent Select Committee on Intelligence (HPSCI) have introduced substantially identical proposals—both called the FISA Reform and Reauthorization Act of 2023—that comprise the anti-reform movement. The HPSCI bill passed out of committee by voice vote on December 7; the Senate bill has been referred to the Senate Judiciary Committee. Some of the key “reforms” touted in these bills include:

• A prohibition on “evidence of a crime only” searches;
• Special protections relating to searches for information about elected U.S. officials and members of Congress;
• Codification of the FBI’s current accountability procedures; and
• Some reforms of the FISA Court process short of Lee-Leahy.

However, as we have noted publicly, the SSCI and HPSCI bills would do more to entrench and even expand surveillance authorities than rein them in. Privacy and civil liberties groups have panned these bills and over one hundred organizations have endorsed the Protect Liberty Act over the HPSCI bill.

As that letter makes clear, the HPSCI and SSCI alternatives to a warrant requirement would do nothing to prevent the vast majority of the documented abuses of Section 702. And merely codifying existing FBI procedures in lieu of real protections is woefully inadequate given evidence suggesting that, even post reforms, the FBI continues to carry out over 200,000 searches a year, including thousands of improper searches.

While both bills’ reforms appear designed to do nothing, other provisions would very clearly expand surveillance authorities in alarming ways. For example, Section 504 of the HPSCI bill vastly expands the number and types of U.S. companies subject to compelled cooperation under Section 702, with huge implications for the scope of this surveillance. Defenders of the bill have argued that the justification for this provision is classified but that they (and the government) interpret the provision “extremely narrow[ly].” There is notably no such limiting language in the text itself.

Both the HPSCI and SSCI bills also have a provision expanding suspicionless searches within Section 702 databases of immigrants and other foreigners seeking to travel to the United States. While the FISA Court has approved a narrower version of this vetting by the NSA, the HPSCI and SSCI bills would effectively excise immigrants as a class from the requirement under Section 702 that searches be reasonably likely to return foreign intelligence information.

And finally, the HPSCI and SSCI bills each eschew the traditional sunset period and would entrench this version of Section 702 for the foreseeable future, extending this authority until 2031 or 2035, respectively. Rather than provide an opportunity for Congress to reevaluate reform efforts in a few years and further calibrate this authority, the HPSCI and SSCI bills would lock these “reforms” in place and stifle any further debate over real reforms.

Looking Forward to 2024

When Congress reconvenes in January, there will likely be substantial pressure to work quickly to reauthorize Section 702, including from the government. It is very likely that this pressure to move quickly will include calls for “compromise” on several of the key areas of reform, such as a warrant requirement or the data broker loophole.

It’s important, then, to not lose sight of this fundamental dichotomy—between real reform bills like the GSRA and the Protect Liberty Act on one hand; and anti-reform bills like those proposed by SSCI and HPSCI. One consistent theme throughout this reauthorization cycle—and those before it—is the government’s intransigence in the face of calls for real reform. The intelligence committee bills would reward that intransigence, despite the well-documented history of abuses of Section 702. Congress can—and must—do better.

---

[1] EPIC has covered Section 702 and the need for reforms in a series of blog posts.

[2] Sen. Rand Paul has also introduced the Fourth Amendment Restoration and Protection Act, which would prohibit the use of FISA authorities (including but not limited to Section 702) as well as EO 12333 surveillance against U.S. persons. See https://www.paul.senate.gov/wp-content/uploads/2023/11/BAG23F354.pdf.