Case Study: The FISA Court — Secret Justice

"The FISC was created to constrain intelligence surveillance. Instead, it became a mechanism for authorizing it." — Adapted from analysis by the Brennan Center for Justice

Overview

The Foreign Intelligence Surveillance Court (FISC, commonly called the FISA Court) occupies a unique position in American governance. Created by Congress in 1978 to provide judicial oversight of intelligence surveillance, the court operates almost entirely in secret. Its judges are appointed by the Chief Justice of the Supreme Court, not confirmed by the Senate. Its proceedings are ex parte — only the government appears. Its opinions were, for decades, entirely classified. And its approval rate for surveillance applications — hovering near 99% for most of its history — has led critics to ask whether it functions as a meaningful check on government power or merely provides a veneer of judicial legitimacy to surveillance operations.

This case study examines the FISA Court's structure, its evolution from a narrow oversight body to a court that has authorized mass surveillance programs, and the ongoing debate about whether secret justice is compatible with democratic governance.

Skills Applied: - Analyzing institutional design and its effects on governance outcomes - Evaluating the tension between secrecy and accountability in democratic institutions - Assessing whether structural reforms can address the FISC's governance deficits - Connecting institutional analysis to broader data governance themes

The Court's Structure

Origins and Design

The FISA Court was created by the Foreign Intelligence Surveillance Act of 1978, enacted in the aftermath of the Church Committee's revelations about warrantless domestic surveillance. Congress intended the court to serve as a judicial check on executive power — ensuring that intelligence agencies could not surveil Americans without some form of independent authorization.

The court's original jurisdiction was narrow: it reviewed applications for electronic surveillance and physical search orders targeting foreign intelligence agents operating within the United States. The standard for approval was that there was probable cause to believe that the target was a "foreign power" or "agent of a foreign power."

Structural features:

  • Eleven judges serve on the FISC, appointed by the Chief Justice of the Supreme Court from among sitting federal district court judges.
  • Non-adversarial proceedings. Only the government presents arguments. The surveillance target is not represented, does not know the proceeding exists, and cannot challenge the application.
  • Classified proceedings and opinions. Sessions are held in a secure facility (a SCIF — Sensitive Compartmented Information Facility). Opinions were historically classified in their entirety.
  • No Senate confirmation. Unlike all other federal judges, FISC judges are appointed by a single individual (the Chief Justice) without Senate confirmation.
  • Rotating service. Judges serve seven-year terms and typically devote only a portion of their time to FISC duties while continuing their regular judicial work.

Expansion Post-9/11

The court's jurisdiction expanded dramatically after September 11, 2001:

  • The USA PATRIOT Act (2001) lowered the standard for FISA orders, expanded the definition of "foreign intelligence information," and created the business records provision (Section 215) that was later used to authorize bulk metadata collection.
  • The FISA Amendments Act (2008) created Section 702, which authorized the collection of communications of non-US persons located outside the United States — without individual court orders for each target. The FISC's role shifted from reviewing individual surveillance applications to approving programmatic surveillance: blanket authorizations for entire categories of collection.

This transformation was pivotal. The FISC was designed to review targeted surveillance requests — "we want to surveil this specific person for this specific reason." Post-9/11, it was asked to approve mass surveillance programs — "we want to collect all telephone metadata for all Americans" and "we want to collect internet communications data from these categories of foreign targets." The court's structure — non-adversarial, secret, deferential — was designed for the former task. It was structurally unsuited to the latter.

The Approval Rate Problem

The Numbers

The FISA Court's approval rate has been the subject of intense scrutiny:

  • From 1979 to 2012, the FISC approved 33,942 surveillance applications and denied 11 — an approval rate of 99.97%.
  • In 2012 (the year before the Snowden revelations), the court approved all 1,856 applications it received, denying none.
  • Post-Snowden disclosures prompted some modifications: the court's annual reports now distinguish between initial applications, modifications, and denials. The numbers still show approval rates above 99%.

What the Numbers Mean

The government argues that the high approval rate reflects the quality of applications, not the weakness of oversight. Intelligence agencies, the argument goes, only submit applications they believe will be approved — poorly supported applications are withdrawn or revised before reaching the court, so the approval rate reflects a filtration process, not judicial passivity.

Critics offer several counter-arguments:

Self-fulfilling deference. If the court almost never says no, intelligence agencies have no incentive to constrain their requests. The approval rate creates a dynamic in which the government pushes for expansive surveillance authority, confident that the court will approve. The court's deference becomes self-reinforcing.

No adversarial testing. In a standard court, the prosecution's arguments are tested by a defense attorney who challenges evidence, raises constitutional objections, and offers alternative interpretations. In the FISC, the government's arguments are uncontested. Research on decision-making consistently shows that decision-makers who hear only one side of an argument are systematically biased toward that side — regardless of their intentions.

Programmatic approval. When the FISC approves a single Section 702 certification, it authorizes the collection of communications affecting potentially millions of people. A single "approval" at the programmatic level is qualitatively different from a single approval of a targeted surveillance order. The approval rate obscures the scale of what each approval authorizes.

Secret Law

The FISC as Lawmaker

One of the most troubling aspects of the FISA Court is that it does not merely apply law — it makes law. The court's classified opinions interpret statutory provisions and constitutional requirements, establishing legal precedents that govern intelligence surveillance. These interpretations have the force of law but are invisible to the public, to most members of Congress, and even to other courts.

The most significant example: in 2006, the FISC approved the NSA's bulk telephone metadata collection program under Section 215 of the Patriot Act. Section 215 authorized the FBI to seek court orders for "tangible things" (business records) that were "relevant to an authorized investigation." The FISC interpreted "relevant" to mean that the entire database of all Americans' telephone metadata was "relevant" to terrorism investigations — because any record might eventually prove useful.

This interpretation was extraordinary. It effectively read the word "relevant" out of the statute, transforming a provision designed for targeted record requests into an authorization for mass data collection. The interpretation was classified. For seven years, the government's official position was that the scope of surveillance under Section 215 was a classified matter — meaning that the legal interpretation authorizing mass surveillance was itself secret.

When the opinion was eventually declassified (post-Snowden), legal scholars expressed shock at the breadth of the interpretation. But for seven years, this secret law had governed the surveillance of hundreds of millions of people.

The Problem of Secret Precedent

The FISC's classified opinions create a body of secret precedent — legal interpretations that bind the intelligence community but are unknown to the public, to legal scholars who might critique them, and to Congress members who might legislate in response. Secret law is antithetical to democratic governance for a basic reason: laws derive their legitimacy from the possibility of public scrutiny and democratic revision. A law that cannot be seen cannot be evaluated, challenged, or changed through democratic processes.

Reform Efforts

Post-Snowden Reforms

The Snowden revelations prompted several reforms to the FISA Court's operations:

Amici curiae. The USA FREEDOM Act (2015) authorized the FISC to appoint independent amici curiae — "friends of the court" — to provide an alternative perspective in cases involving "a novel or significant interpretation of the law." This was the most significant structural reform, introducing a limited adversarial element into proceedings that had been entirely one-sided.

However, the amicus provision has limitations. Appointment is at the court's discretion, not mandatory. The amici serve as legal advisors, not advocates for surveillance targets. They do not have independent investigative authority. And they are drawn from a pre-approved panel — not from the broader civil liberties community.

Declassification of significant opinions. The court began publishing declassified versions of significant opinions, providing some transparency into its legal reasoning. But declassification is selective — the court and the government jointly determine which opinions are "significant," and redaction can remove the most operationally consequential details.

Annual reporting. The FISC now publishes annual transparency reports with statistics on applications received, approved, modified, and denied. These reports provide quantitative data but limited qualitative insight into the court's reasoning.

Proposals for Deeper Reform

Scholars and advocates have proposed more fundamental reforms:

  • Mandatory adversarial representation. A permanent "public advocate" within the FISC who represents the interests of surveillance targets in every proceeding — not just those the court selects.
  • Senate confirmation of FISC judges. Bringing FISC appointments under the same democratic accountability as other federal judicial appointments.
  • Systematic declassification. Automatic declassification of FISC opinions after a fixed period (e.g., five years), with narrow exceptions for genuinely sensitive operational details.
  • Appellate review. Strengthening the Foreign Intelligence Surveillance Court of Review (FISCR) and allowing civil liberties organizations to appeal FISC rulings.

None of these reforms has been fully implemented. The intelligence community has resisted each, arguing that greater transparency would compromise national security.

The Governance Paradox

The FISA Court embodies a governance paradox: it was created to constrain intelligence agencies, but its structure makes it more likely to enable them. The non-adversarial process ensures that the government's arguments are never challenged. The secrecy ensures that errors of judgment are never exposed. The lifetime appointment of the appointing authority (the Chief Justice) ensures that the composition of the court cannot be influenced by democratic elections.

Dr. Adeyemi posed the question to her class directly: "Is a secret court that almost never says no providing oversight, or providing authorization? If the answer is authorization, then the FISC is not a check on power — it is a mechanism that launders unchecked power through the appearance of judicial process."

Discussion Questions

  1. The government argues that the FISA Court's high approval rate reflects the quality of applications. Under what conditions would you find this argument persuasive? What evidence would you need to see?

  2. The amicus curiae reform introduced a limited adversarial element into FISC proceedings. Is a limited adversarial element sufficient, or is mandatory adversarial representation necessary? What are the risks and benefits of each approach?

  3. Compare the FISA Court's governance model with the governance of algorithmic systems discussed in Chapters 16-17. Both involve complex technical decisions made behind closed doors with limited external review. What structural features do effective oversight mechanisms need, whether the subject is surveillance law or algorithmic fairness?

  4. Should classified legal interpretations ever be permissible in a democracy? If so, under what conditions? If not, how should national security interests be balanced against the democratic requirement for legal transparency?

  5. Eli argues that the FISA Court and the Smart City sensors in his neighborhood share a common structure: decisions made about people without their knowledge, consent, or participation, justified by claims of public benefit. Do you find this parallel valid? Why or why not?

Further Investigation

  • Read selected declassified FISA Court opinions (available through the Federation of American Scientists' Intelligence Resource Program or the Office of the Director of National Intelligence).
  • Research the Privacy and Civil Liberties Oversight Board (PCLOB) and its assessments of the Section 215 and Section 702 programs.
  • Compare the UK's Investigatory Powers Tribunal with the FISA Court. Which provides more meaningful oversight?