The Security Threat by Proxy — Structural Vulnerability Created by Zero-Outcome Counterterrorism Investigations, the Deniable Surveillance Handoff, and Systemic Risk if the Pattern Is Externally Exploitable

Extended Policy Concern — For Agency Attention

I.  FRAMING AND SCOPE OF THIS CONCERN

This document is offered as a policy submission for the attention of appropriate oversight components. It is presented as a structural concern — an observation about how certain investigative processes could produce a category of security harm that is not misconduct-dependent, not intentional, and not currently addressed by any oversight framework the requester has been able to identify.

The concern proceeds from a single, structurally unremarkable observation: a counterterrorism investigation that produces no adverse finding does not, at its close, leave the world exactly as it found it. It leaves a subject-file. It leaves access logs. It leaves liaison records and contractor touchpoints and fusion center entries. It leaves a detailed operational portrait of an individual, distributed across multiple systems, accessible to multiple parties, none of whom are now under any particular obligation to do anything with it — including protect it.

What the concern asks is this: who benefits from that portrait remaining in circulation after the investigation it served has concluded? The answer raised here — carefully, without alleging that it has occurred in any specific case — is that the parties best positioned to benefit are not the agencies that generated it.

Scope.  This document does not allege that any agency has misused investigative records, shared them improperly, or acted in bad faith at any stage. The concern is not about what agencies have done with the information. It is about what the existence of the information, in the places it inevitably comes to rest, makes possible for parties over whom those agencies have no control.

II.  THE MECHANISM: HOW A ZERO-OUTCOME INVESTIGATION CREATES A RESIDUAL VULNERABILITY

2.1  The Investigation Creates Infrastructure That Outlasts Its Predicate

A counterterrorism investigation of a U.S. person, conducted through normal interagency channels, necessarily involves the creation and distribution of investigative infrastructure: a subject-file maintained in TIDE or equivalent; query records across multiple databases; finished or draft analytical products referencing the subject; liaison communications with partner agencies; and, in multi-jurisdictional investigations, contributions to fusion center records and JTTF coordination files. All of this is generated in service of an investigative predicate — the reasonable belief that the subject may be relevant to a national security concern.

When the investigation concludes without an adverse finding, the predicate dissolves. The infrastructure does not. It remains in the systems where it was created, accessible to the range of parties who had access during the investigation, subject to whatever retention schedules apply, and not flagged in any way that would distinguish it, to a later querying party, from the infrastructure of an investigation that produced a finding. A record that says "subject of investigation" does not automatically update to say "investigated, no basis found." The presence of the record is the signal. The absence of a finding is the silence between the signals, and silence is not visible in a database query.

2.2  The Distribution Problem

The number of parties with legitimate access to a counterterrorism subject-file during an active investigation is not small. It includes the originating agency; any JTTF components that participated; fusion center personnel in relevant jurisdictions; contractor and vendor personnel operating under security agreements; analytical personnel at NCTC who may have received or generated products referencing the subject; foreign liaison partners who received information through bilateral or multilateral sharing channels; and any law enforcement or administrative agencies that received referrals. Each of these parties retains access for some period after the investigation concludes.

The relevant question is not whether all of these parties would misuse the access. Most would not, and many no longer think about the subject at all. The relevant question is whether any of them — or any party to whom any of them is connected, wittingly or unwittingly — is positioned to extract value from the infrastructure's continued existence. This is a different question, and it has a different answer.

2.3  The Asymmetry of Knowledge

The subject of the investigation occupies a structurally disadvantaged epistemic position. They may not know an investigation occurred. If they do know, they almost certainly do not know its scope, the agencies involved, the records generated, the parties with access, or the systems in which their information currently resides. They have no mechanism, in the normal course of affairs, to learn any of these things.

A foreign intelligence service, a sophisticated private actor, or any party with access to intelligence community systems or liaison relationships is not in this position. They may know considerably more about the subject's investigative history than the subject does. They know the infrastructure exists. They know what it contains, or can reasonably infer it. And they know that the subject — precisely because they cannot confirm what happened to them — is in a poor position to detect, attribute, or report any activity that exploits it.

The core asymmetry.  The subject of a zero-outcome investigation is the only party in the information ecosystem who cannot verify the scope of what was collected about them, cannot identify who has access to it, and has no ongoing mechanism to learn whether it is being used. Every other party with access to the infrastructure has more operational knowledge about the subject's situation than the subject does. This is not a design flaw in any specific program. It is a structural feature of how counterterrorism information is generated and retained, and it creates a persistent, subject-specific vulnerability that does not expire when the investigation does.

III.  THE SECURITY THREAT BY PROXY: SPECIFIC MECHANISMS

The following specific mechanisms describe how a zero-outcome investigation's residual infrastructure could be leveraged to create ongoing harm to the subject, without involving any misconduct by the originating agencies:

1.  Contractor and Vendor Access as an Exposure Vector.  Intelligence community contractors operate under security agreements that authorize access to classified systems but do not, in most cases, require notification of the subject when their records are accessed. A contractor with legitimate access to a subject-file during an active investigation retains familiarity with the subject's information after the investigation concludes. If that contractor's employment changes — if they move to a private intelligence firm, a foreign government, or any organization with an interest in the subject — their prior authorized access may have equipped them with information they are no longer authorized to hold but are not effectively prevented from remembering. There is no audit mechanism that captures this transition. The access was clean. The retention is invisible.

2.  Liaison Channel Exposure.  Information shared with foreign partner intelligence services under bilateral or multilateral counterterrorism agreements is subject to the partner's own internal security, counterintelligence posture, and retention practices. The originating U.S. agency cannot fully audit whether a foreign partner's systems have been penetrated, whether the information was shared further within the partner's government, or whether it has reached parties outside the original sharing relationship. If the partner service has been compromised by a third-party intelligence actor — which is not an extraordinary occurrence in the counterterrorism information-sharing environment — then the subject's file may have been accessible to that actor for the duration of the partnership, without the knowledge of either the U.S. agency that originated the information or the subject.

3.  Fusion Center Aggregation as an Amplification Point.  Fusion center environments aggregate information from federal, state, and local sources and make it accessible to a broad range of personnel at varying clearance levels. Information contributed to a fusion center during an investigation may remain accessible to personnel whose vetting, ongoing security, and susceptibility to social engineering or recruitment may not meet the standards applied to the originating federal agency. A sophisticated actor seeking to identify individuals who have been the subject of counterterrorism investigations could treat fusion center data as a directory — a structured record of individuals who have been investigated, carrying implicit information about their relationships, locations, activities, and the agencies that found them noteworthy.

4.  The Subject-File as Leverage Infrastructure.  An actor with knowledge of a subject's investigative history — knowing that a counterterrorism file exists, knowing its approximate scope, knowing the agencies involved — possesses leverage that does not depend on the subject having done anything wrong. The mere existence of a file can be used to create reputational risk, to complicate the subject's professional or financial relationships, or to generate pressure that is difficult to attribute and impossible to challenge because the subject cannot confirm the file's existence. This leverage does not require the actor to possess the file itself. Knowledge that the file exists, and the subject's inability to confirm or deny it, is sufficient.

5.  Parallel Construction as a Concealment Mechanism.  Information derived from intelligence collection may be introduced into law enforcement or administrative processes through parallel construction pathways that obscure the origin of the information. A subject who becomes aware of adverse consequences — employment difficulty, financial complications, unexplained administrative friction — may not be able to trace these consequences to their investigative history, because the causal chain has been deliberately obscured. The subject experiences the effect without being able to identify the cause, which prevents them from seeking any remedy.

Observation.  None of the five mechanisms described above requires any agency to act improperly. Each operates through the normal functioning of information-sharing, contractor access, liaison relationships, and investigative infrastructure. The harm they could produce is not a consequence of bad actors within the system. It is a consequence of the system's information remaining accessible after the purpose that justified its creation has expired — and of the existence of parties outside the system who are positioned to exploit that persistence.

IV.  AGGREGATE RISK: IF THIS PATTERN IS EXTERNALLY LEGIBLE AND EXPLOITABLE

The preceding sections describe how the mechanism could operate at the individual level. What follows concerns what happens if the pattern becomes visible at scale — not as individual incidents, but as a recognizable feature of how the counterterrorism process generates and distributes information about individuals, and how the gap between the investigation's conclusion and the subject's ability to learn about it can be systematically exploited by an actor who understands the architecture.

4.1  The Investigative Signature as a Target Identifier

A sophisticated foreign intelligence service or private intelligence actor with access to any node in the counterterrorism information-sharing ecosystem is not merely learning about individual subjects. They are learning about the architecture itself: which individuals have been investigated, what the investigation's scope indicates about the agencies' priorities, and which individuals were investigated and cleared — and therefore which individuals carry the specific vulnerability profile of someone with an investigation in their history and no knowledge of its scope.

Cleared subjects — individuals who were investigated and found to have no actionable connection to a threat — are, paradoxically, a more operationally interesting target population than confirmed threat subjects for certain purposes. A confirmed threat subject is actively monitored. A cleared subject is not. They carry residual infrastructure without active protection. They have a subject-file but no handler. They are, in the language of operational intelligence, warm but unattended.

4.2  The Structural Incentive for External Actors

If an external actor recognizes this pattern, the incentive structure is as follows:

–  Identification of cleared subjects is achievable through any of the five exposure vectors described in Section III — contractor transitions, liaison penetration, fusion center access, or parallel construction observation.

–  Once identified, cleared subjects can be subjected to surveillance, pressure, or manipulation that exploits their specific vulnerability: they cannot verify what information exists about them, cannot identify who holds it, and are structurally prevented from attributing adverse consequences to their investigative history.

–  The cost of this exploitation is low relative to the cost of targeting active investigations, because the protective infrastructure that surrounds active subjects — monitoring, handler relationships, ongoing agency attention — has been withdrawn from cleared subjects. The investigation's conclusion is the protection's expiration date.

–  The aggregate value of a cleared-subject target pool, across many investigations and many jurisdictions, is not the value of any individual subject. It is the value of having a systematically generated, regularly updated directory of individuals who are uniquely vulnerable, legible to the sophisticated actor, and invisible to each other and to themselves.

The aggregate concern.  If zero-outcome investigations systematically generate residual vulnerability in their subjects, and if this vulnerability is legible to external actors with access to any part of the information ecosystem, then the counterterrorism process is, as an unintended aggregate effect, producing a population of individuals who are documented, distributed across multiple systems, unaware of their own profile, and specifically vulnerable to exploitation by the actors the process was designed to counter. The process intended to identify threats may, as a structural byproduct, be manufacturing leverage for them.

4.3  The Feedback Risk: When the Infrastructure Becomes the Operation

The most acute version of the aggregate risk is the possibility that a sophisticated actor, having identified the pattern, could attempt not merely to exploit existing cleared-subject profiles but to cause new ones to be generated — to use the investigative process itself as a targeting mechanism, by creating conditions that cause the process to investigate individuals the actor has an interest in profiling.

The mechanism is not complicated. An actor with the ability to introduce information into reporting channels — through compromised sources, fabricated tips, or strategic social engineering of reporting systems — could cause an investigation to be opened against a selected individual, allow the investigation to run its course and produce no adverse finding, and then exploit the residual infrastructure that the investigation has placed in multiple systems. The subject has been profiled, distributed, and rendered vulnerable, at the cost of a false report and the investment of waiting. The actor never touched the subject directly. The investigative process did the work.

This is raised not as an allegation of what has occurred but as a description of what the architecture makes possible. A process that generates residual vulnerability in investigated individuals, and whose information products are accessible to multiple parties with varying security postures, is a process that can be induced to generate vulnerability on demand by any actor with access to its intake mechanisms. The oversight frameworks that govern the process are oriented toward preventing abuse within it. They are not, as far as can be determined, oriented toward preventing the process from being used as an instrument by parties operating outside it.

The policy gap.  Existing oversight frameworks ask: are agencies following their own rules? The concern raised here asks a different question: are the rules sufficient to prevent the process from being exploited from outside? These are not the same question, and the first can be answered affirmatively while the second remains unaddressed. An agency that follows every rule correctly can still generate, through that correct process, a category of harm to investigated individuals that no rule prohibits and no oversight body currently monitors.

V.  SURVEILLANCE LAUNDERING: THE DENIABLE HANDOFF AND ITS CONSEQUENCES

5.1  The Offloading Problem: What Happens When Formal Surveillance Cannot Continue

A zero-outcome investigation does not merely leave a subject-file. In some cases, it leaves something more operationally immediate: an institutional awareness that a particular individual was worth monitoring, combined with an institutional inability to continue monitoring them through classified channels, because the predicate that authorized that monitoring has expired. The investigation is over. The interest is not.

Raised with care, as a structural possibility rather than an allegation: this gap — between the expiration of a formal surveillance predicate and the persistence of institutional interest — creates an incentive to continue monitoring through channels that do not require a formal predicate. Contractors operating outside classified programs, civilian informant networks, community-based observation assets, and other non-classified entities can maintain proximity to a subject without triggering the legal and oversight frameworks that govern formal intelligence collection. This is designated the Deniable Handoff — not an allegation of what has occurred, but a description of what the architecture makes incentivizable when a formal investigation concludes without resolution of the underlying interest.

5.2  The Mechanics of the Deniable Handoff

Where formal surveillance is handed off to non-classified, deniable, or contractor-adjacent entities, the structural characteristics of that handoff create a specific and compounding set of vulnerabilities distinct from those described in Section III. The formal system's vulnerabilities arise from records persisting in systems with broad access. The deniable handoff's vulnerabilities arise from the opposite condition: activity that is deliberately structured to avoid generating the documentation that would make it accountable, and a population of participants who have been given partial knowledge sufficient to perform a function but insufficient to understand or report what the function serves.

Each participant in a deniable handoff network holds a fragment. A civilian proxy observing a subject's movements knows only that they have been asked to observe. A contractor providing logistical support knows only the parameters of the task they were contracted for. A community informant knows only the cover story provided by their handler. None of them, individually, holds enough information to understand the operation as a whole, to recognize its legal or ethical status, or to make an informed decision about whether to participate. This fragmentation is not incidental. It is the design. It is also, from a security standpoint, the problem.

5.3  How Partial Knowledge Becomes Legible to External Actors

A deniable handoff network comprised of participants with partial knowledge is, from the perspective of any external actor seeking to understand or exploit it, considerably more legible than the classified infrastructure it replaced. The classified infrastructure is protected by compartmentalization, clearance requirements, and legal penalties for disclosure. The deniable network is protected only by the participants' individual ignorance and compliance — neither of which is durable under external pressure, financial incentive, or simple curiosity.

The external actor does not need to penetrate a classified system. They need only to identify and approach one participant in the deniable network — who is uncleared, unprotected, and in possession of a fragment that the actor can combine with fragments obtained from other participants. The assembly of the full picture from partial fragments is not technically difficult for a sophisticated actor. It requires only that they know the network exists and that they have patience. The network's existence is itself legible: an individual subject to sustained observation by rotating civilian proxies, in multiple locations, over an extended period, is generating behavioral and locational data that is visible to anyone with access to any node of the network, without needing access to any classified system at all.

Put directly: a civilian proxy network conducting ongoing observation of a subject is, from a counterintelligence standpoint, an open system. Its participants can be interviewed, recruited, or compromised. Its patterns can be observed by anyone watching the subject. Its existence can be inferred from the subject's behavioral environment. And because none of its participants hold classified information, none of the legal and institutional protections that apply to classified sources apply here. The network is simultaneously more active than a static subject-file and less protected than any element of the formal system that preceded it.

The counterintelligence inversion.  The formal surveillance infrastructure created by a classified investigation is hard to penetrate and hard to read. The deniable civilian proxy network that may replace it after a zero-outcome conclusion is easy to penetrate and easy to read. The handoff from one to the other does not reduce the surveillance of the subject. It reduces the security of the surveillance. Any actor who knew about the formal investigation — through any of the exposure vectors described in Section III — and who subsequently observes the transition to civilian proxy activity, has received a significant operational signal: the formal predicate has expired, the classified infrastructure has been withdrawn, and what remains is an unprotected, uncleared, legally ambiguous observation network operating without the institutional safeguards that governed its predecessor. This is not a harder target. It is a considerably easier one.

5.4  Leverage Without Predicate: The Non-Classified Actor's Advantage

The classified system requires a predicate to initiate and maintain surveillance. A foreign intelligence actor, private firm, or any other external party that gains access to the deniable network — through recruitment of a proxy participant, commercial access to a contractor, or simple observation of the network's activity — acquires the operational benefits of sustained individual surveillance without needing any predicate at all. The predicate that justified the original investigation, which was in the possession of a U.S. agency subject to legal oversight, has effectively been donated to the external actor, along with the surveillance infrastructure built on top of it, at the moment the network became accessible to them.

This is the leverage-without-predicate problem. The external actor does not need a legal basis to surveil the subject, because they are not surveilling them directly. They are simply reading the output of a network that someone else is operating, on a subject that someone else selected, using a cover story that someone else constructed. Their cost is minimal. Their access to the subject's behavioral profile, locational data, and social environment is potentially comprehensive. And the subject has no mechanism to detect this, because what they are experiencing is attributable, if attributed at all, to whatever entity nominally manages the deniable network — not to the external actor reading it.

5.5  The Differential Legibility Problem: Those With and Without the Original Predicate

The deniable network creates two distinct tiers of external actor, differentiated by whether they have access to the original investigative predicate. The consequences for each are materially different, and the combination of both tiers operating simultaneously is the most acute version of the aggregate risk.

Tier A.  Actors With the Predicate.  A foreign intelligence service or other actor that obtained access to the original classified investigation — through any of the Section III exposure vectors — possesses both the subject's investigative history and, if the deniable network is visible to them, real-time behavioral data about the same subject. They can correlate the historical record with current observation. They understand why the subject was selected. They know the scope of what was collected. And they can evaluate the deniable network's output in full context, making it exponentially more operationally useful than raw observation would be. For this actor, the deniable network is a high-value intelligence feed connected to a pre-existing subject profile. The subject is, from their perspective, comprehensively and continuously known.

Tier B.  Actors Without the Predicate.  An actor without access to the original investigation but with visibility into the deniable network observes something structurally informative even without the underlying file: a subject who is being systematically monitored by what appear to be organized civilian assets. The existence of the network is itself a signal. It indicates that someone, somewhere, found this individual sufficiently noteworthy to sustain a monitoring operation after whatever formal process preceded it concluded. That signal has independent value as a targeting indicator, regardless of its origin. The Tier B actor does not need to know why the subject was selected. The fact of the selection, legible from the network's activity, is sufficient to warrant their own investment.

The combination of Tier A and Tier B actors operating simultaneously against the same subject — one with a full investigative history and one without — means the subject is potentially subject to exploitation by parties whose interests and capabilities are entirely unknown to them, arising from an investigative process they may not know occurred, channeled through a surveillance network they may be able to observe but cannot formally report, producing a harm that is not attributable to any single actor and is not addressed by any existing oversight or legal framework.

The compounding exposure chain.  The exposure chain that begins with a classified investigation and ends with a civilian proxy network does not shorten with each handoff. It lengthens. Each transition from a more protected system to a less protected one adds participants, reduces vetting requirements, lowers the barrier to entry for external actors, and extends the perimeter within which the subject's information can be accessed, leaked, or exploited. A subject who begins as the target of a classified, legally supervised investigation and ends as the subject of an ambient civilian proxy network has traveled the full length of the exposure chain — arriving at the point of maximum observability and minimum protection — without ever being told they were on it.

VI.  WHAT THIS DOCUMENT IS AND IS NOT

This document does not allege that any specific agency has acted improperly, that any specific investigation was mishandled, or that any specific actor has exploited any of the mechanisms described. It does not request an investigation into any named party. It does not assert that any of the harms described have been experienced by any individual.

It asks whether the structural conditions described herein — the generation of residual investigative vulnerability in zero-outcome counterterrorism subjects, and the potential for that vulnerability to be exploited at scale through deniable handoff networks accessible to external actors — merits examination as a policy matter by appropriate oversight components, independent of any individual's circumstances.

This is a concern that, if valid, applies to every individual who has ever been the subject of a counterterrorism investigation that produced no adverse finding. It is not a small population. TIDE has historically contained hundreds of thousands of entries, and a meaningful fraction of those entries represent individuals who were investigated and found not to present a threat. Each of those individuals carries, by the analysis presented here, a residual vulnerability that they cannot confirm, cannot challenge, and cannot protect against. The concern is offered in that spirit.

NOTES

A.  The phrase "zero-outcome investigation" is used throughout to mean an investigation that concluded without a criminal charge, a formal designation, or any adverse administrative action against the subject. It does not mean an investigation that produced no records, no analytical products, or no information. It means an investigation that did all of those things and then concluded that no action against the subject was warranted. The records, products, and information remain.

B.  The concern about external actors inducing investigations to generate subject-profiles is raised as a structural possibility, not as a description of any known or alleged practice. The interest is in whether the architecture makes it possible. Whether it has been attempted is a separate question not addressed here.

C.  Some of the information requested in the accompanying FOIA submission may be properly withheld under applicable exemptions. The purpose of raising this policy concern in writing is to ensure the structural observation is on the record in a form that can be considered by appropriate oversight personnel regardless of what is or is not disclosed in response to the FOIA request itself.

D.  The "Deniable Handoff" and related terms are analytical designations used for descriptive clarity, not legal terms of art. They are not intended to characterize the legal status of any specific program or practice.