How FISA Section 702 Became a Bipartisan Surveillance Bargain

There's a particular kind of political vertigo that sets in when the party that spent two years warning about authoritarian overreach turns around and hands the sitting president one of the most powerful domestic surveillance tools ever written into law. That vertigo is available in abundance right now, as a contingent of House Democrats appear poised to join Speaker Mike Johnson in renewing Section 702 of the Foreign Intelligence Surveillance Act — a provision whose scope, critics argue, has never been fully reckoned with by the public it affects.

As The Intercept reported on March 23, 2026, Democrats crossing the aisle to support renewal are potentially providing the margin Johnson needs to push the reauthorization through. Rep. Jamie Raskin, one of the more vocal skeptics in his own caucus, framed the situation with characteristic bluntness: "What could go wrong with that?"

The question is rhetorical. But the structural answer — the one that explains not just what could go wrong but why institutions keep arriving at this same juncture — is worth unpacking carefully.

What Is FISA Section 702, and Why Does It Keep Coming Back?

Section 702 of the Foreign Intelligence Surveillance Act, enacted as part of the FISA Amendments Act of 2008, authorizes the U.S. government to conduct warrantless surveillance of foreign nationals located outside the United States. On paper, that sounds narrow. In practice, it is extraordinarily broad.

The mechanism works like this: because foreign communications routinely pass through U.S.-based servers and infrastructure, the National Security Agency (NSA) can compel American telecommunications companies to hand over communications involving targeted foreign individuals — even when those communications include American citizens on the other end. Those American communications, captured incidentally, are then stored in searchable databases. The FBI, NSA, and CIA can query those databases for American names without obtaining a warrant.

According to the Office of the Director of National Intelligence, there were approximately 246,073 non-U.S. persons targeted under Section 702 in 2023 alone. The number of American communications swept up in the process is not disclosed publicly — a gap that civil liberties advocates have consistently highlighted as a fundamental transparency failure.

The law is subject to periodic reauthorization, which means Congress must affirmatively vote to keep it alive. That expiration mechanism was designed as an accountability check. In practice, it has functioned as a recurring political theater piece, in which lawmakers debate civil liberties concerns, attach modest reforms, and ultimately renew the authority anyway.

The Pattern Beneath the Politics

What makes this moment structurally interesting isn't the surveillance debate itself — that debate has been ongoing since at least the 2013 Snowden disclosures. What's interesting is the pattern by which institutional actors repeatedly choose continuity over constraint, regardless of which party holds power.

I've written before on PatternThink about how institutions preserve their own surveillance and control architectures even when the stated justification for those architectures shifts. Section 702 is a near-perfect case study.

Consider the timeline:

The pattern is consistent: alarm, debate, modest concessions, renewal. Each cycle slightly normalizes the previous one. The Overton window on warrantless surveillance of Americans moves incrementally, not in leaps.

Every reauthorization of Section 702 since 2013 has occurred with bipartisan support, regardless of which party controlled the White House — demonstrating that surveillance authority transcends partisan alignment once embedded in the national security apparatus.

Why Are Democrats Crossing the Aisle Now?

This is the question that has generated the most commentary, and the answers being offered in Washington tend to be tactical: some Democrats represent swing districts and don't want to appear soft on national security; others genuinely believe the intelligence value of 702 outweighs its civil liberties costs; still others may be negotiating for reforms they believe they can extract as part of a deal.

But there's a deeper structural explanation that tactical analysis misses.

Intelligence bureaucracies are extraordinarily effective at generating dependency. Once a surveillance tool is woven into the operational fabric of the FBI, NSA, and CIA — once case officers, analysts, and field agents are trained on it, once prosecutions are built around it, once the infrastructure for collecting and querying the data is fully built out — the institutional pressure to preserve that tool becomes enormous, regardless of who sits in the Oval Office.

Lawmakers who receive classified briefings on 702-derived intelligence are shown the wins. They are shown the terrorism plots disrupted, the foreign agents identified, the threats neutralized. They are rarely shown — because it is structurally difficult to show — the full scope of incidental collection of American communications, the queries run on political figures, journalists, or activists, or the cases where the tool was misused.

The FBI conducted an estimated 3.4 million queries of Section 702-collected data involving U.S. persons in a single year, according to a 2023 report from the Privacy and Civil Liberties Oversight Board (PCLOB) — a figure that stunned even some longtime surveillance reform advocates.

Democrats who crossed the aisle in previous reauthorization cycles often justified it on the grounds that they could extract reforms. The reforms that have actually been codified — ending "about" collection, adding some query logging requirements — are real but limited. The core architecture of warrantless access to American communications remains intact.

What the Trump Variable Changes

Previous Section 702 reauthorization debates happened under administrations — Obama's, Trump's first term, Biden's — that, whatever their other characteristics, operated within recognizable norms of intelligence oversight. The DOJ Inspector General, the FISA Court, and the PCLOB functioned as at least partial checks on abuse.

The current political environment is categorically different in one important structural respect: the Trump administration has been explicit about its willingness to use the instruments of federal law enforcement against political opponents, journalists, and perceived enemies. The President and his allies have explicitly framed political adversaries as threats to national security. That framing, combined with a surveillance tool that allows warrantless queries of Americans' communications, creates a structural risk that is qualitatively different from previous eras.

Jamie Raskin's pointed question — "What could go wrong with that?" — is not merely partisan sniping. It reflects a genuine recognition that tools built for one purpose do not stay bounded to that purpose when the people wielding them have broader ambitions.

The civil liberties concern here is not hypothetical. The FISA Court's own opinions, portions of which have been declassified, document a history of FBI compliance failures with the rules governing 702 queries. In a 2022 declassified FISA Court opinion, the court found the FBI had improperly queried Section 702 data for information about January 6th defendants, Black Lives Matter protesters, and a sitting U.S. senator. That was under an administration not openly hostile to those groups. The implications under one that is should not require a long chain of inference.

The Reform Proposals on the Table

The current debate is not simply "renew or don't renew." Several reform proposals have been circulating, ranging from marginal to genuinely structural:

The structural problem with all of these proposals — including the strongest, the warrant requirement — is that they attempt to regulate behavior inside an architecture that was designed to minimize judicial oversight. The warrant requirement is the most meaningful reform available, but intelligence agencies have consistently argued, with some success, that adding a warrant requirement would effectively make the tool unusable.

That argument deserves scrutiny. The claim that a judicial warrant requirement for searching American communications would cripple national security is not self-evidently true — it is a policy argument advanced by agencies with institutional interests in preserving frictionless access. The Fourth Amendment's warrant requirement has never been understood by courts to be incompatible with national security investigations; the question is whether the FISA Court can be adapted to process such requests efficiently, not whether judicial oversight is categorically impossible.

The Institutional Logic of Bipartisan Surveillance

There is a reason this story keeps repeating across administrations, across party transitions, across years of civil liberties advocacy and public outcry. It is not that lawmakers are uniquely corrupt or cowardly, though individual examples of both exist. It is that surveillance authority, once established, becomes structurally self-reinforcing.

Intelligence agencies develop doctrine around available tools. Prosecutors learn to build cases that incorporate surveillance-derived evidence. Oversight bodies become acculturated to the scope of collection. The "normal" baseline shifts with each reauthorization cycle. By the time any particular reauthorization comes up, the question is never really "should we have this power?" The question is always "should we take it away?" — a much harder political and institutional lift.

This is the deeper pattern that Section 702 illustrates: democratic accountability mechanisms work well at preventing the acquisition of new powers; they work poorly at constraining or eliminating powers that have already been normalized through repeated use.

The Democrats currently considering a vote for renewal are not, for the most part, enthusiasts for domestic surveillance. They are operating within an institutional logic that makes "yes, with reforms" almost always easier than "no." That logic is not partisan. It is structural. And understanding it structurally is prerequisite to any serious attempt to change it.

For readers interested in the broader architecture of how institutions resist accountability, the patterns of institutional self-preservation explored here on PatternThink are directly relevant to understanding why surveillance reform movements consistently underperform their apparent political support.

What to Watch Going Forward

Several indicators will determine how this particular cycle resolves:

1. The reform package: Whether any warrant requirement survives into the final bill will be the clearest signal of whether this reauthorization represents a meaningful shift or pure continuity.

2. The vote breakdown: The specific names of Democrats who vote yes matter. Members of the Judiciary Committee, who have been most directly exposed to evidence of FBI misuse, will send the clearest signal about institutional knowledge versus institutional loyalty.

3. The FISA Court's posture: The Court's willingness to publicly flag compliance failures — as it has done in recent years — is one of the few structural checks that has shown some independence. Whether that posture survives a more politically hostile environment is an open question.

4. Executive branch signaling: The Trump administration's explicit statements about how it intends to use 702 authority will be politically significant, both for the vote and for any subsequent legal challenges.

The underlying structural question — whether a democratic legislature can meaningfully constrain a surveillance architecture it has normalized over nearly two decades — will not be resolved by this reauthorization cycle. But this cycle will reveal something important about how far that normalization has progressed.

Raskin's question deserves a real answer, not a rhetorical one. The answer, structurally, is: quite a lot. The more important question is whether that answer changes the vote.

Jared Clark is the founder of PatternThink, where he writes about the hidden structural patterns that shape institutions, organizations, and human systems.

Last updated: 2026-03-30