The Watcher Turns Around: How Intelligence Agencies Rediscover Their Domestic Appetite

There is a detail buried in the administrative history of the Roman Empire that deserves more attention than it typically receives. The frumentarii — originally grain-supply officers, logistical functionaries tasked with keeping the legions fed — were gradually converted by successive emperors into something else entirely: an internal surveillance and enforcement network that monitored senators, tracked dissidents, and reported private conversations back to the palace. The transformation took generations. No single emperor sat down and decided to build a secret police. The institution simply followed the gravity of its own utility.

That story is not Roman. It is human. And the five thousand years of political history available to us suggest that any society wishing to avoid it must contend with something more durable than any individual administration's intentions: the institutional logic of the intelligence apparatus itself.

The Tool Finds Its Own Uses

Intelligence organizations are created with a specific external threat in mind. The threat is real. The organization is built. And then — almost always — something happens that political scientists sometimes call mission creep and that historians, with the benefit of distance, tend to call something more honest: mission completion.

The logic runs like this. An agency develops collection capabilities, analytical methods, informant networks, and technical infrastructure at enormous expense. Those capabilities do not distinguish between foreign enemies and domestic ones. The people running the agency are not, as a rule, sociopaths; they are bureaucrats responding to incentives. When a domestic political problem presents itself — a labor movement that seems suspiciously organized, a religious minority that communicates in unfamiliar ways, a journalist whose sources are embarrassingly accurate — the tools are already there. Using them feels, to the people inside the institution, like efficiency.

The Stasi of East Germany is the most thoroughly documented example in the modern record. At its peak, the Ministry for State Security maintained a ratio of roughly one informant for every sixty-three East German citizens — a penetration rate that no foreign intelligence service has ever matched against an adversary. It is worth remembering that the Stasi was not created to surveil East Germans. It was created, in the standard formulation of the Cold War era, to protect the state from Western subversion. The domestic surveillance apparatus grew because the tools were available and because internal dissent, once it was defined as foreign-inspired, became indistinguishable from foreign threat.

The American Record

The United States has its own version of this history, and it is neither as extreme as the Stasi nor as clean as the civics textbook.

COINTELPRO, the FBI's Counter Intelligence Program active from 1956 to 1971, was formally aimed at the Communist Party of the United States — an organization with genuine, documented ties to Moscow. But the program expanded with a momentum that its architects may not have fully anticipated. By the late 1960s, it encompassed the Southern Christian Leadership Conference, the National Association for the Advancement of Colored People, the Socialist Workers Party, and the women's liberation movement. The FBI sent a forged letter to the wife of a civil rights leader suggesting her husband was having an affair. It attempted to convince a Black nationalist organization that a rival group had placed a contract on its leadership. These were not foreign intelligence operations. They were domestic political operations conducted with the infrastructure of a foreign intelligence service.

The post-9/11 expansion of NSA collection authorities, revealed in detail by the Snowden disclosures of 2013, followed a structurally similar path. The legal authorities constructed to intercept foreign terrorist communications were broad enough — and the technical architecture of modern telecommunications is interconnected enough — that domestic communications were collected at scale. The Foreign Intelligence Surveillance Court, created by Congress in 1978 specifically to provide judicial oversight of foreign intelligence gathering, had by 2013 issued secret rulings that authorized bulk collection programs its drafters almost certainly never envisioned.

None of this requires a conspiracy theory. It requires only an understanding of institutional behavior across time.

Oversight as the Only Counterweight

History does offer a partial answer to the surveillance creep problem, and it is not a reassuring one: robust, adversarial oversight from an institution that is genuinely independent of the agency it oversees. The Church Committee of 1975 — the Senate Select Committee that investigated FBI and CIA abuses — produced reforms that constrained domestic intelligence activity for a generation. Those reforms held, more or less, until the pressure of a genuine external threat provided the political environment for their quiet erosion.

The pattern suggests that oversight is not a permanent achievement. It is a recurring contest. The Byzantine Empire maintained a complex of competing intelligence services precisely because emperors understood that any single service, left unchecked, would eventually serve itself. The Venetian Republic's Council of Ten, feared as it was, operated under a system of rotating membership and mandatory review that kept it from calcifying into a private power center — for a while.

The American intelligence community today operates under more formal legal constraints than at any point in its history. It also possesses more technical capability than at any point in its history. Whether the first fact meaningfully limits the second is a question that the five-thousand-year record of similar institutions does not answer with optimism.

What the Record Tells Us

Human psychology has not changed. The official who genuinely believes a domestic critic is a foreign asset is not lying — he may simply be using the tools available to him to confirm what his institutional position has already suggested is true. The emperor who dispatched the frumentarii to monitor a senator was not, in his own understanding, building a secret police. He was protecting the empire.

This is what makes surveillance creep so difficult to address through normal political means. It does not feel like overreach to the people doing it. It feels like diligence.

The only reliable check the historical record identifies is external, structural, and sustained: an oversight mechanism with real investigative power, real independence, and real political will to use both. Where those conditions have existed, the apparatus has been constrained. Where they have eroded — as they always eventually do, under the pressure of genuine external threat or simple bureaucratic attrition — the watcher has, with remarkable consistency, turned around.