The Government Can Buy Your Data Without a Warrant. Congress Has 18 Days to Decide If That's Okay.
What's Happening
Section 702 of the Foreign Intelligence Surveillance Act expires on April 20. This reauthorization cycle comes with a question that didn't exist last time: should the federal government be allowed to buy Americans' personal data from commercial brokers, feed it through AI systems, and conduct pattern-of-life analysis without a warrant?
Right now, the answer is yes.
Federal agencies can purchase location data, browsing histories, and communications metadata from data brokers instead of requesting a warrant. AI has changed what's possible with that data. What once required teams of analysts and months of work can now be done in seconds: behavioral profiles, association mapping, movement prediction. The intelligence community says this capability has produced real results against real threats. Civil liberties advocates say it's an end-run around the Fourth Amendment.
Two bipartisan bills, the Government Surveillance Reform Act (Wyden-Lee) and the SAFE Act (Lee-Durbin), would require warrants for government data purchases. A coalition of 130 civil society organizations and 17 state attorneys general support closing the loophole. Intelligence professionals warn that warrant requirements could slow time-sensitive operations against active threats.
The same commercial data brokers selling to U.S. agencies also sell to foreign governments, private intelligence firms, and anyone else willing to pay. The $13.4 billion the Department of War requested for AI and autonomous systems in FY2026 is built partly on commercial data streams that adversaries can access through the same market.
Congress has 18 days.
Here's My Two Cents
I've been in rooms where both sides of this argument are operational, not theoretical. The intelligence value of commercial data acquisition is real. I'm not going to pretend otherwise. And the constitutional concerns about warrantless mass data collection are real. I'm not going to pretend about that either.
What frustrates me is that these aren't contradictory positions. The hawk is right that this capability fills gaps traditional collection can't. The reformer is right that the Fourth Amendment wasn't designed with a terms-of-service workaround. A well-designed warrant framework preserves the capability while putting it on legal ground that can survive the next administration, the next scandal, and the next court challenge. That's not a compromise. That's just good governance.
But here's the part that actually makes me angry. We have bipartisan bills. We have bipartisan sponsors. We have a problem both sides acknowledge. And Congress still might not act because this issue doesn't break cleanly along party lines, and issues that don't break cleanly don't get fundraising emails. So instead the taxpayer funds a surveillance capability that exists in legal limbo, built on a commercial supply chain our adversaries can shop at, with no statutory framework and no democratic accountability.
I'd tell my hawk friends: a warrant requirement doesn't kill the capability. It puts it on ground that won't collapse when the next whistleblower shows up on 60 Minutes. That protects the operators, not just the public. I'd tell my reformer friends: perfect shouldn't be the enemy of good. The GSRA isn't everything you want, but it closes the most dangerous loophole currently in play.
Congress has until April 20. The bipartisan framework exists. The legislative text exists. The coalition exists. The question is whether Congress will do its job or punt again and leave taxpayers funding something nobody voted for.
Anna R. Dudley writes on national security, intelligence policy, and the places where hawks and reformers need to find each other. Weekly Update is a weekly series. Subscribe at annardudley.substack.com.