A Lapse That Changed Nothing — Yet
On June 5, 2026, the Senate voted 47-52 against advancing a reauthorization of Section 702 of the Foreign Intelligence Surveillance Act, with seven Republicans — Josh Hawley, Mike Lee, Rand Paul, Eric Schmitt, Rick Scott, Tommy Tuberville, and John Kennedy — joining nearly every Democrat to block the motion (Roll Call; CBS News). The House declined to pass a short-term patch, and the statute lapsed at midnight on June 12 — the first time Congress has let this authority expire since it was created in 2008 (EFF).
Here is the detail that should reframe how everyone covers this story: almost nothing changed on the ground. Section 702 collection runs on annual certifications approved by the Foreign Intelligence Surveillance Court, and the FISA Amendments Act's own transition language lets certifications already in effect at sunset continue until they expire on their own terms. The certifications approved in March 2026 run to roughly March 2027, so the NSA, FBI, and CIA keep operating under the current rules for another eight months regardless of what Congress does next (FISC public filings). The lapse is a procedural fact, not an operational one — Congress has bought itself real time to legislate carefully rather than under a deadline gun, and it should use it.
The Case for the Program, Stated Fairly
Section 702 supporters have a genuine argument, and it shouldn't be waved away. The authority — created by a 2008 amendment to FISA and reauthorized most recently by the Reforming Intelligence and Securing America Act (RISAA), Public Law 118-49, in April 2024 — lets agencies compel U.S. communications providers to hand over the communications of specifically targeted foreigners located abroad, without a judicial warrant for each target (Congressional Record, May 2, 2024). Congress has repeatedly found value in this collection for tracking terrorist plots, foreign cyber intrusions, and state-sponsored espionage; requiring an individualized warrant before targeting a foreign national on foreign soil would be a novel and arguably unworkable constraint that no other democracy imposes on its own signals-intelligence services. Intelligence officials have long warned that any lapse risks a collection gap precisely when adversaries are adapting their tradecraft. That warning deserves to be taken seriously, even if — as this lapse shows — the practical gap turns out to be closer to zero than advertised.
Where the Program Actually Goes Wrong
The honest problem with Section 702 isn't the targeting of foreigners abroad — it's what happens after their communications land in a government database that also, incidentally, contains Americans'. The FBI and other agencies query that database using U.S.-person search terms without a warrant, a practice civil liberties groups call "backdoor searches." The Brennan Center for Justice, drawing on the government's own compliance reporting, documents that the FBI ran over 278,000 noncompliant queries in a single year, 2022, and that agents have searched the database for members of Congress, journalists, a sitting federal judge, protesters, and roughly 19,000 political donors (Brennan Center). Those are not edge cases; they are the predictable result of giving investigators warrantless access to a haystack that happens to contain millions of Americans' private messages. RISAA's 2024 reforms added supervisory sign-off requirements for U.S.-person queries, but they stopped short of the one fix privacy advocates and a bipartisan bloc of legislators — the same coalition that sank the June 5 vote — have pushed for years: a judicial warrant before the FBI can read an American's communications collected under this program.
"No warrant to protect Americans? No FISA," Senator Mike Lee said in explaining his vote — a soundbite, but not an unfair summary of the coalition's position.
What Congress Should Do With the Runway It Has
The pro-innovation, pro-security answer here is not abolition. Section 702 collection against genuine foreign targets is a narrower, more legally disciplined program than the bulk domestic metadata collection Edward Snowden exposed in 2013, and gutting it entirely would remove a tool that has demonstrably disrupted real plots. But the current expiration, with its built-in nine-month operational runway, is the best opportunity in a decade to attach a warrant requirement for U.S.-person queries — with the emergency, consent, and cyber-defense carve-outs that reform bills have already modeled — without anyone being able to claim national security is being held hostage to the negotiation. Polling cited by the Brennan Center puts support for a warrant requirement at roughly 76% among Americans across party lines; this is not a fringe position forcing itself into the debate, it is where the public already is.
Congress should also resist the temptation to smuggle unrelated riders into what should be a narrow, high-stakes surveillance-reform bill. The lapse bought lawmakers until March 2027 to get this right. Squandering that runway on unrelated fights, or on another photo-finish short-term patch, would be the real failure here — not the expiration itself.