**Links**: [Blogger](https://bryantmcgill.blogspot.com/2026/04/civilian-spycraft.html) | [Substack](https://bryantmcgill.substack.com/p/washingtons-privateers-of-intelligence) | [Obsidian](https://bryantmcgill.xyz/articles/The+Coming+Market+for+Civilian+Spycraft) | Medium | Wordpress | [Soundcloud 🎧](https://soundcloud.com/bryantmcgill/civilian-spycraft)
**Don't be shy... you could be a private spy.**
*Spying is no longer a curiosity reserved for trench-coated professionals or signal-intercepting agencies; it is the ambient operating condition of contemporary American life, in which ordinary citizens, corporate security officers, contractors, OSINT investigators, journalists, platform operators, threat-intelligence vendors, mercenary-spyware firms, fusion-center participants, and AI agents now occupy overlapping rungs of the same intelligence ladder. What follows is an expanded historical and structural synthesis tracing how the American republic was born through civilian intelligence and privateering against empire, how that same template was industrialized by Allan Pinkerton's detective agency from Civil War espionage into corporate strikebreaking and ultimately into a Swedish-owned multinational risk-management subsidiary still operating today, how the Obama-era Nationwide Suspicious Activity Reporting Initiative quietly inducted hundreds of thousands of line officers and private-sector security personnel into a federal reporting architecture, how the USD 14–21 billion 2026 OSINT market sits atop a USD 250-billion-plus data broker substrate that the intelligence community now consumes through ODNI's centralized commercially-available-information portal, how Palantir's USD 145-million-plus ImmigrationOS contract has made the contractor-as-sovereign-function pattern operationally explicit, how Black Cube and the broader private-spy industry continue to shape elections and reputational warfare from Iceland to Slovenia, how mercenary spyware vendors target civil society at industrial scale, how 2025 brought a simultaneous revival of cyber letters of marque (the Schweikert bill) and cartel letters of marque (the Lee-Burchett bills) recovering the Article I, Section 8 dormant clause that has not been activated since the War of 1812, and how agentic AI OSINT systems—GeoSeer, MiroFish, Blue Helix, OWASP SocialOSINTAgent, the agentic SOC consensus at RSA 2026, the Anthropic-Accenture autonomous security workflow—have collapsed the marginal cost of continuous civilian surveillance toward zero. The thread running through these layers is a single ontological observation: toolhood is a politically convenient under-description of a relation tending toward twinship, and the privateer, the Pinkerton, the suspicious-activity reporter, the data broker, the contractor, the mercenary-spyware operator, the citizen analyst, and the autonomous agent are all variations on the same recurrent figure—the state-adjacent collector operating without the state's badge—now reconstituted at planetary scale and machine speed. The piece closes on what civic intelligence literacy must mean for citizens who wish to remain free agents inside a substrate that has learned to read them continuously.*
## I. The Citizen as Sensor, the Sensor as State
The CIA's recent essay, **"Five Lessons in Spying, Courtesy of George Washington,"** is useful not merely because it reminds us that Washington was a sophisticated intelligence commander, but because it unintentionally points toward a much larger historical recurrence: intelligence in America has never belonged only to formal agencies. It has always depended on **border figures**—surveyors, merchants, couriers, sailors, tavern keepers, printers, innkeepers, farmers, ferrymen, ship captains, contractors, informants, and civilians who could move through ordinary life without appearing to be part of the state. Washington's lessons—know the adversary, cultivate soft skills, collect current intelligence, understand blind spots, and build circles of trust—are presented as the early grammar of American espionage. But beneath that grammar is a more unsettling pattern: the American intelligence tradition begins not with a sealed bureaucracy but with **distributed civilian participation**, improvised under existential pressure and later formalized into institutions, contracts, and platforms.
The question now is whether that old pattern is returning under digital conditions, not as the Culper Ring but as a privatized, platform-mediated, **AI-augmented market for private citizen tradecraft and spycraft**, sitting atop a mass-market surveillance economy that already touches almost every American without their meaningful consent. The Open Source Intelligence sector alone is variously estimated at between **roughly USD 14 billion and 21 billion in 2026**, with major analysts projecting growth toward USD 43–133 billion by the early to mid-2030s, depending on definitional scope, and the data broker industry that feeds it pulled in **more than USD 250 billion in 2022** by some industry counts. ([Mordor Intelligence][1], [Global Market Insights][2], [Brennan Center][3]) These are not boutique numbers. They describe an industrial infrastructure within which civilianized intelligence collection has already become a normal commercial activity, contracted by governments, corporations, journalists, investigators, advocacy networks, political campaigns, and private clients alike.
The thesis of this piece is therefore not nostalgic and not alarmist. It is structural. **America was born through civilian intelligence and privateering against empire**; it has continually re-domesticated those tools through corporate detective agencies, fusion-center reporting architectures, contractor-driven analytic capacity, and now agentic AI systems that can plan and execute multi-step open-source investigations with limited human oversight. The privateer of the eighteenth century needed a ship, crew, cannon, insurance, investors, and a letter of marque. The **private intelligence actor of the twenty-first century** may need only access, tooling, legal ambiguity, narrative skill, and an audience or client. The vessel has become the platform. The prize has become behavioral insight, reputational leverage, strategic advantage, coercive exposure, or monetizable intelligence. The sea has changed, but the temptation has not. Public purpose and private appetite are again learning to wear the same flag.
## II. Washington's Civilian Intelligence Atmosphere
AMC's *TURN: Washington's Spies* understood this better than many conventional patriotic histories. The series is not reliable as literal historiography in every detail, and historians have rightly noted that it compresses chronology, invents dramatic overlays, and sometimes turns historical ambiguity into character melodrama. But its deeper value is symbolic. *TURN* shows espionage not as glamorous technique but as **social fracture**. Abraham Woodhull is not James Bond in homespun clothing; he is a compromised civilian living inside a divided community, where loyalty, family, commerce, fear, opportunism, ideology, and survival all occupy the same room. Anna Strong, Benjamin Tallmadge, Caleb Brewster, Robert Townsend, John André, Benedict Arnold, and even the terrifying Robert Rogers become less like isolated characters than nodes in an **intelligence ecology**. Every dinner table is a listening post. Every friendship is a possible channel. Every market exchange carries strategic residue. Every ordinary object may become a carrier of meaning. That is the real overlap between Washington's world and ours: **intelligence emerges wherever ordinary life becomes strategically legible**. Mount Vernon describes the historical Culper Ring as a network created under Washington and Major Benjamin Tallmadge to collect information from British-occupied New York; *TURN* dramatizes that network as a civilian atmosphere, a place where war moves through households before it moves through armies. ([Mount Vernon][4])
Washington's intelligence genius was therefore not simply that he "used spies." Many commanders used scouts, informants, deception, rumor, intercepted correspondence, and local guides. His deeper achievement was that he recognized intelligence as a **living interface between geography, psychology, timing, trust, and civilian cover**. As a young envoy in the Ohio Country, he learned that terrain was not just land but intention made spatial. As a military commander, he learned that current intelligence could turn high ground into survival or ignorance into catastrophe. Around Braddock, he learned that institutional confidence can become an analytic blind spot. During the Revolution, he learned that a circle of trust had to be small enough to survive betrayal yet distributed enough to see what formal command could not. These are not merely "spy lessons." They are lessons in **informational sovereignty**: how a vulnerable system senses danger, filters deception, preserves secrecy, and acts before superior force closes around it. ([cia.gov][5])
The Culper Ring's world was dangerous because ordinary social life concealed divided sovereignty. A farmer might be a patriot, loyalist, opportunist, informant, smuggler, courier, reluctant collaborator, or terrified neutral. A dinner conversation could become intelligence. A romantic attachment could become vulnerability. A tavern could become a data exchange. A ferry route could become an intelligence artery. The twenty-first-century equivalent is not the candlelit inn but the **platform profile, the encrypted chat, the location trail, the freelance investigation, the corporate risk dashboard, the leak archive, the AI-summarized social graph, and the private report sold to a client**. The village has become a database. The tavern has become the timeline. The courier has become the API. The privateer has become the contractor, the OSINT investigator, the patriotic hacker, the data broker, the threat-intelligence vendor, the mercenary spyware firm, and the citizen analyst. The Culper Ring did not have a SaaS dashboard, but its underlying architecture—**civilian sensors plus discreet routing plus trusted analysis plus protected identity**—is exactly the architecture being rebuilt at planetary scale today.
## III. Privateers, Not Pirates: The Letter of Marque as Constitutional Architecture
That is where the privateer analogy becomes essential, because the same Founders who institutionalized civilian intelligence also institutionalized **state-sanctioned private violence at sea**. The American Revolution was not fought only by Continental soldiers or official naval vessels. Private shipowners could receive legal authorization—**letters of marque or privateer commissions**—to attack enemy commerce under state authority. The National Park Service explains the distinction clearly: a Letter of Marque authorized armed merchant ships to challenge enemy vessels during commercial voyages, while a Privateer Commission applied to vessels whose primary purpose was disrupting enemy shipping. In both cases, a private actor became an extension of sovereign violence, operating with legal sanction but also with profit incentives, ambiguity, and risk. ([NPS][6])
Privateering was not an ornamental sideshow. It was a **major instrument of asymmetric war**. During the Revolution, private vessels gave a weak maritime power reach against a global empire; during the War of 1812, Congress issued roughly 1,100 letters of marque, and American privateers captured at least 1,200 British vessels, compared with roughly 250 seizures by the U.S. Navy. By some accounts, the 1776 Continental Navy fielded a mere 64 ships, while privateer vessels under license numbered roughly **1,697**—a force ratio that helps explain why letters of marque are now being seriously revived in policy imagination after a 200-year dormancy. ([American Battlefield Trust][7], [TIME][8]) Privateering was cheap, scalable, deniable in spirit if not always in law, and emotionally attractive because it allowed ordinary citizens to imagine themselves as participants in national defense. But it also fused patriotism with prize-taking. It rewarded initiative while inviting abuse. It blurred the distinction between public purpose and private appetite. It created a zone in which the state could **outsource risk while private actors absorbed danger and pursued reward**.
The architecture matters because the United States never signed the 1856 Paris Declaration that abolished privateering among major powers. The dormant clause—Article I, Section 8, Clause 11—remains constitutionally live. Congress can still grant letters of marque and reprisal. ([Center for Maritime Strategy][9]) When that historical reality is put alongside the privatization trends in contemporary intelligence and cyber operations, what looks like a fringe novelty—revived privateering bills—is better understood as **a constitutional architecture being recovered to legitimize what is already happening informally** in the OSINT, contractor, threat-intelligence, mercenary-spyware, and crisis-intelligence markets.
## IV. The Missing Link: Pinkerton, the Eye That Never Sleeps, and the Industrialization of Private Intelligence
Between Washington's Culper Ring and the present-day OSINT vendor, there is a historical bridge that most contemporary discussions skip: the **Pinkerton National Detective Agency**, founded around 1850 by Allan Pinkerton, a Scottish-immigrant cooper turned Chicago deputy sheriff. The Pinkerton story is the single most important nineteenth-century proof that civilian intelligence in America does not stay civilian—it congeals into a private firm, gets contracted by the state during emergencies, then turns its capacities back on the domestic population for industrial clients. Read carefully, Pinkerton is the genetic ancestor of every modern private spy company, every corporate threat-intelligence team, every rented investigator, and every "trust and safety" risk operation that quietly maps activists, journalists, or labor organizers on behalf of paying customers.
The arc is almost too neat to be accidental. Pinkerton was an abolitionist who hosted Underground Railroad fugitives, gave John Brown money before Harpers Ferry, and built America's first private detective agency under the slogan **"We never sleep,"** an unblinking eye whose iconography produced the term *private eye*. ([Wikipedia: Pinkerton][10], [Britannica][11]) When the Civil War came, he served the Union as **Major E.J. Allen**, running the Union Intelligence Service for General McClellan, claiming to have foiled the Baltimore Plot to assassinate Lincoln, and providing—by the harsh judgment of military historians like T. Harry Williams—**"the poorest intelligence service any general ever had,"** primarily because Pinkerton's interrogation methods systematically inflated Confederate troop numbers and immobilized McClellan into siege warfare. ([Wikipedia: Allan Pinkerton][12]) That is itself the first lesson: **a private firm's strategic incentives may not align with the analytic discipline that sovereign decisions require**, even when patriotism is sincere and the contractor competent. The contractor profits from the contract; the consumer of the intelligence pays the strategic cost.
After the war, the state contracted Pinkertons to perform functions that did not yet have public agencies. In **1871, Congress appropriated USD 50,000** (roughly USD 1.3 million in present terms) to the new Department of Justice for "detection and prosecution," but the sum was insufficient to staff an internal unit, so DOJ contracted the Pinkertons to do the work. ([Wikipedia: Pinkerton][10]) This is the hinge moment, easily lost in a footnote: the **federal government, in its first attempt to build a national investigative capacity, used a private firm as substitute infrastructure** because the state's own organs were not yet capable. The Pinkerton template—**private firm fills public capacity gap, becomes embedded, accumulates data and tradecraft no public agency yet possesses**—is precisely the template of Palantir's relationship with ICE in 2025 and 2026, of Recorded Future's relationship with intelligence customers, of Babel Street's relationship with defense and law enforcement, of LexisNexis Risk Solutions' relationship with DHS and ICE. The grammar has not changed. Only the substrate has.
What Pinkerton did next is the part nineteenth-century industrialists understood and twenty-first century citizens forget: he turned his **state-honed intelligence capability inward, against domestic populations, for corporate clients**. His sons William and Robert ran detective and protection branches; protection in this period meant **infiltrating labor unions and breaking strikes**. The agency was hired against coal, iron, lumber, and railroad workers across multiple states. The Library of Congress notes that "Pinkerton guards and agents gained notoriety as strikebreakers." ([Library of Congress][13]) During the **Homestead Strike of 1892**, Henry Clay Frick of Carnegie Steel called in 300 Pinkerton agents to break the strike at the Pittsburgh-area mill. The resulting firefight killed roughly nine to eleven steelworkers and three Pinkertons, depending on which contemporary count one credits, and produced sufficient national outrage to prompt the **Anti-Pinkerton Act of 1893**, which prohibits the federal government from contracting "an individual employed by the Pinkerton Detective Agency, or similar organization." ([Wikipedia: Pinkerton][10]) That statute remains on the books, although its scope has been substantially narrowed by interpretation, and it stands today as one of the few American laws that ever attempted to **legislate a structural separation between sovereign function and private detective agency**—precisely the boundary now being eroded again.
The hunt for Jesse and Frank James between roughly 1875 and the early 1880s is an even darker echo. While trying to capture the brothers, Pinkertons firebombed the family farmhouse in Missouri in January 1875, severely injuring Zerelda Samuel—the brothers' mother—and killing their eight-year-old half-brother Archie. William Pinkerton later admitted that the pursuit had become **"a war of extinction."** ([All That's Interesting][14]) The mid-nineteenth-century private detective agency, contracted both by railroads and indirectly by adjacent state interests, had drifted into something a lawful republic could not formally authorize but could quietly tolerate. The drift was not from competence to incompetence—it was from **investigation toward extralegal kinetic action**, from intelligence work toward something closer to private war.
This Pinkertonian arc is a precise rehearsal for everything that followed. After the La Follette Committee hearings of 1937 publicly exposed corporate labor espionage, Pinkerton diversified. By the 1960s the word "detective" disappeared from its letterhead. The firm rebranded as **personal security and risk management**, was acquired in 1999 by the Swedish security conglomerate Securitas AB for USD 384 million, and now operates as **Pinkerton Consulting & Investigations** under the parent's umbrella, focused on threat intelligence, risk management, executive protection, and active-shooter response. The former Pinkerton Government Services arm operates as **Securitas Critical Infrastructure Services**. ([Wikipedia: Pinkerton][10]) The eye that never sleeps did not close. It was absorbed into a multinational consulting business whose modern services—threat intelligence, executive protection, infrastructure security—are the legitimate, professional surface of an industry whose deeper history includes private war against strikers, extralegal kinetic operations against outlaw families, and contracted substitution for public investigative functions before public investigative agencies existed.
The reason the Pinkerton story matters now is that it falsifies the comforting belief that private intelligence is a recent aberration produced by post-9/11 contractor expansion or by Silicon Valley's data appetite. **Private intelligence has always been a constitutive feature of the American state**, not an exception to it. The republic was built with civilian Culper sensors, maritime privateers, and within a generation of independence, a private detective industry willing to take government contracts, corporate contracts, or both, often simultaneously. What is new in our moment is not the existence of the privateer or the Pinkerton. It is the **scale, speed, automation, and ambient invisibility** at which their successors now operate, and the increasing willingness of legislators to formalize once again, under updated names, what the Pinkertons did in the gap between Washington's civilian sensors and the eventual professional federal bureaucracy.
## V. The Suspicious Activity Reporting Architecture: Civilianized Collection by Design
The softer phrase **"public reporting"** does not fully name the structure. When private citizens, private security officers, infrastructure operators, corporate partners, and local intermediaries are trained to notice pre-incident indicators, document suspicious behavior, route observations into fusion centers, and feed reports into federal information-sharing systems, the functional category is **civilianized intelligence collection**. In plain language, yes: this is spying. It may be lawful, patriotic, useful, and sometimes necessary; it may also be overbroad, abusive, socially corrosive, and constitutionally hazardous. But once the state trains non-state actors to gather observations on persons, activities, movements, associations, photography, elicitation, surveillance, infrastructure interest, or anomalous behavior and then sends those observations into counterterrorism or law-enforcement intelligence repositories, **the public has crossed from ordinary vigilance into an intelligence ecology**.
The Obama-era example is important precisely because it was not a cartoon villain program. It was forward-leaning, civic, post-9/11, interoperability-minded, and wrapped in the language of prevention, community trust, privacy, civil rights, and homeland security. That is what made it historically significant. The **Nationwide Suspicious Activity Reporting Initiative (NSI)** was not merely a slogan campaign. The Government Accountability Office reported in 2013 that DOJ had largely implemented NSI among fusion centers, that the program used technical mechanisms such as **Shared Spaces servers** and the FBI's **eGuardian** system, and that DOJ had trained executives at 77 of 78 fusion centers, roughly **2,000 fusion center analysts, and about 290,000 of 800,000 line officers**. ([GAO][15]) This was not a handful of concerned neighbors. It was a national reporting architecture with formal training, databases, vetting processes, fusion centers, and federal-state-local interoperability.
The documented architecture also supports the recollection of multiple input pathways. Official NSI call-to-action materials describe several entry paths into the **NSI SAR Data Repository**, including the eGuardian User Interface, the SAR Vetting Tool, and records-management systems connected to the repository. The same materials state that tips or leads could originate from law-enforcement officers, public-safety agencies, private-sector partners, or citizens before being reviewed, analyzed, and potentially submitted into the repository for sharing with NSI participants. That is a pipeline. One can debate policy justification, abuse rates, and necessity, but the machinery itself is not ambiguous: **private observations were meant to become standardized intelligence objects**.
The public-facing slogan, **"If You See Something, Say Something,"** becomes more serious when placed inside that machinery. The Congressional Research Service notes that DHS launched the national campaign in July 2010 to raise awareness of indicators of terrorism and terrorism-related crime, and tied the campaign to NSI as a way of emphasizing community awareness. CRS also identifies community-outreach efforts that sought to connect private citizens to the SAR process and build awareness among public, private-sector, and law-enforcement actors. ([Every CRS Report][16]) The slogan was the surface grammar; the deeper structure was a **reporting-and-fusion architecture** that took perception itself as raw input.
The private-sector training language is even more direct. The NSI private-sector security course states that its goal is to train private-sector security officers in their role of **observing and reporting suspicious activity** under the NSI SAR Functional Standard, and that the course teaches suspicious behaviors associated with pre-incident terrorism activities and how and where to report them. Once a private actor is trained to observe behavior according to an intelligence standard and report it into government-linked channels, that actor has been **partially inducted into state-adjacent collection**. The badge may be absent, the oath may be absent, the title may be absent, but the function is present. The ACLU's *Gill v. DOJ* challenge framed SAR as a "vast expansion" of the domestic intelligence network, arguing that the program encouraged reporting of lawful activity, including First Amendment-protected activity, into counterterrorism databases. Whether one accepts every part of the ACLU critique or not, it correctly identifies the underlying structural issue: **the reporting threshold can become elastic enough that normal behavior becomes suspicious once viewed through a pre-crime lens**. ([ACLU][17])
The Obama-era system was therefore **forward-leaning but flawed**. It anticipated the modern reality that threats are distributed, that attacks may be preceded by observable behaviors, that local actors often see fragments before federal agencies do, and that prevention requires networked sensing rather than centralized command alone. But it also normalized the civic habit of watching one another through state-defined suspicion categories. It taught people to convert neighborly perception into reportable intelligence. It carried the **soft authoritarian risk** built into every participatory security system: the citizen becomes not merely protected by the state, but invited to become a low-level perceptual extension of it.
The InfraGard layer shows the same pattern in a more explicitly public-private form. The FBI's own 2012 information-sharing report describes InfraGard as a partnership among the FBI, state and local law enforcement, academic institutions, businesses, and other organizations dedicated to protecting critical infrastructure by sharing information on cyber and physical threats and vulnerabilities. The report states that InfraGard's goal is an ongoing dialogue with the FBI for investigative and intelligence-gathering purposes, that the FBI provides alerts through secure email and a secure website, and that members share information with the FBI and each other through secure websites, bulletin boards, email, list servers, and meetings. This is not a rumor structure; it is **a documented private-sector intelligence-sharing architecture**, later described in reporting as an 80,000-plus member portal. The exact figure—20,000, 80,000, 200,000, or more—depends on which layer is being counted: trained officers, private-sector members, readership networks, fusion-center partners, community participants, or reportable citizen inputs. The more important point is that the architecture is real, distributed, scalable, and now sits beneath the platform-mediated reality of 2026.
## VI. The Open-Source Substrate: When the Public Web Becomes Intelligence Material
This is not a speculative fringe development; it is already visible in official doctrine. In 2024, the Office of the Director of National Intelligence and CIA released the **IC Open Source Intelligence Strategy 2024–2026**, defining OSINT as intelligence derived from publicly or commercially available information that addresses intelligence priorities, requirements, or gaps. The strategy explicitly treats OSINT as vital to the intelligence mission, calls for expanded acquisition and sharing of open-source data, integrated collection management, innovation, partnerships with industry and academia, and the development of a **next-generation OSINT workforce and tradecraft**. ([ODNI][18]) That phrase—**next-generation OSINT workforce and tradecraft**—is the hinge. Once the open world itself becomes an intelligence substrate, the line between agency analyst, contractor, journalist, investigator, researcher, platform operator, corporate risk team, patriotic volunteer, and private citizen becomes structurally unstable.
The positive version is already celebrated. **Bellingcat** and similar open-source investigative communities have shown that civilians can use publicly available material to document war crimes, track weapons movements, verify events, identify perpetrators, and challenge official narratives. The modern civic investigator can sometimes see what governments deny or overlook. Journalists use OSINT to expose corruption, locate atrocities, and hold powerful actors accountable. The International Center for Journalists describes OSINT as the gathering and analysis of publicly available information from social media, databases, and government records, especially valuable where information is sparse, controlled, or censored. This is **citizen tradecraft as democratic audit function**—not espionage against the public, but disciplined observation in defense of public truth. ([ICFJ][19])
The market scale tells the rest of the story. By 2026, the OSINT market by varying estimates sits between roughly **USD 14 billion and USD 21 billion**, with major analyst houses projecting growth toward USD 43–133 billion by the early to mid-2030s, depending on definition. North America commands roughly 36–47 percent of installations; cloud-based deployments dominate at around two-thirds of share; private-sector OSINT usage has reached approximately 38 percent of all deployments globally. Vendors like Google, Palantir, Recorded Future, Babel Street, Thales, Digimind, and dozens of specialist firms now offer text analytics, social-media analytics, geospatial intelligence, dark-web monitoring, identity resolution, and adverse-media screening at scales unimaginable to any nation-state intelligence service of fifty years ago. ([Mordor Intelligence][1], [Global Market Insights][2], [Industry Research][20]) **The public sphere has become machine-readable, and the machines doing the reading are mostly commercial.**
But the darker version is growing beside it. The same social graph that helps expose a war crime can help target a dissident. The same commercial data market that helps map humanitarian crises can reveal religious attendance, protest participation, romantic relationships, immigration vulnerability, military family patterns, or journalist-source proximity. The same private intelligence skill set that can expose corruption can be hired to suppress allegations, intimidate witnesses, profile reporters, or manipulate institutions. The OSINT layer is **morally ambidextrous by structural design**, and the same toolchain that supports a Bellingcat investigation supports a corporate suppression operation when the client check clears.
## VII. The Data Broker Surveillance Economy: The Real Substrate Beneath OSINT
Behind the celebrated layer of disciplined civilian investigation sits a vastly larger and almost entirely invisible commercial substrate: the **data broker industry**, which collects, infers, aggregates, packages, and sells information about Americans on a continuous basis with negligible federal regulation. EPIC and other privacy advocates document **thousands of brokers** operating in the United States, building profiles on essentially every American adult—name, address, telephone, email, age, marital status, children, education, profession, income, political preferences, real-estate ownership, mobile advertising IDs, location histories, purchase histories, financial relationships, religious-attendance inferences, mental-health inferences, military-family signals, and far more. ([EPIC][21], [Privacy Rights Clearinghouse][22]) The Brennan Center estimates the industry pulled in more than **USD 250 billion in 2022**. ([Brennan Center][3]) These are not buried back-office operations; they are the load-bearing economy of the consumer internet, embedded in adtech, fraud detection, identity verification, credit decisioning, marketing, political targeting, and now law enforcement and immigration enforcement.
The Duke University data brokerage research project, led by Justin Sherman, has shown that **sensitive non-public data on active-duty military members, veterans, and their families can be purchased from U.S. brokers for as little as USD 0.12 per record**, often without serious customer verification—one broker offered to skip identity checks if the buyer paid by wire instead of credit card. ([Duke Tech Policy][23], [Lawfare][24]) The same datasets contain health, financial, and religious-practice information. If a foreign intelligence service can buy individually identified data on American servicemembers for pennies per record, the **operational distinction between commercial marketing and adversarial targeting collapses at the database layer**. In December 2024, the Consumer Financial Protection Bureau issued a proposed rulemaking, *Protecting Americans from Harmful Data Broker Practices*, attempting to bring brokers under the Fair Credit Reporting Act framework Congress originally enacted in 1970 specifically to constrain the then-emerging consumer data surveillance industry. ([Federal Register][25], [CFPB][26]) Whether the proposed rule survives political transition will determine whether the United States accepts this substrate as governable infrastructure or as a permanent unregulated second-tier economy.
The cleanest demonstration that this infrastructure has fused with state intelligence is **The Intercept's May 2025 reporting** on ODNI's plan to build a centralized "one-stop shop" portal for **commercially available information (CAI)**—a unified system through which intelligence community elements can ingest brokered data, including location data derived from mobile ads, that would previously have required a court order. The portal will eventually allow analysis with large language models and "sentiment analysis," and will include data deemed by ODNI as highly sensitive, that which can be "misused to cause substantial harm, embarrassment, and inconvenience to U.S. persons." ([The Intercept][27]) Senator Ron Wyden, a long-standing critic of the **Fourth Amendment end-run** that CAI represents, framed the concern bluntly: policies are one thing, but actual government use of bulk commercial data about Americans is the operational reality. The phrase "just grab all of it, we'll find something to do with it" captures the inversion: rather than collecting on identified targets within constitutional limits, the system collects everything on everyone within commercial markets and **filters retroactively under analytic categories selected by whoever holds the portal credentials**. This is, at structural level, the inverse of warrant-based collection.
LexisNexis Risk Solutions and Thomson Reuters CLEAR, the dominant commercial aggregators, hold detailed profiles on virtually every American adult—**78 billion records on individuals**, by some industry counts—and CLEAR is used by more than **3,400 law enforcement agencies**. ([State of Surveillance][28]) When DHS pays LexisNexis roughly USD 9.75 million to provide "identity verification services," what is actually procured is access to a privately built mass-surveillance corpus that no public agency could lawfully assemble itself. The Pinkertonian template, reread through this lens, is unmistakable: **a private firm fills a capacity gap the state could not lawfully fill on its own, becomes structurally indispensable, and accumulates data and tradecraft no public agency yet possesses.** The eye that never sleeps no longer wears a badge or a logo; it wears the brand of a data analytics firm and a security clearance.
## VIII. Palantir, ImmigrationOS, and the Contractor as Sovereign Function
If the data broker economy is the substrate, **Palantir Technologies is the most concrete contemporary example of the contractor-as-sovereign-function** pattern operating at scale. Palantir has held an Investigative Case Management (ICM) contract with ICE since 2014; in April 2025 it was awarded a USD 30 million contract to upgrade ICM into a new platform called **ImmigrationOS** (Immigration Lifecycle Operating System), with prototypes scheduled for delivery in September 2025 and the contract running through September 2027. The platform's three stated functions are **targeting and enforcement prioritization**, **self-deportation tracking** with "near real-time visibility," and **immigration lifecycle management** from identification through removal. ICE's own contract justification described the procurement as "urgent and compelling" and named Palantir as the "only source" capable of delivering it in time, citing the company's "deep institutional knowledge of ICE operations" and active ingestion of data from multiple federal agencies. ([American Immigration Council][29], [Axios][30]) By late 2025, the ICM-ImmigrationOS contract envelope had ballooned to over USD 145 million, and Palantir had received more than **USD 900 million in federal contracts** since the start of the second Trump administration. ([ACLU][31])
The architecture matters more than the dollar amounts. Palantir's ICM and ImmigrationOS pull data from passport records, Social Security files, IRS tax data, license-plate reader streams, and field-collected device extracts uploaded by ICE agents. ICE training documents reportedly instruct agents on how to extract and upload data from phones seized at the border or during arrest. By January 2026, 404 Media had described "**ELITE**," a Palantir application that ICE uses to identify neighborhoods to raid based on aggregated profile criteria. ([ACLU][31], [Immigration Policy Tracking][32]) In December 2025, Fortune reported a new contract between Palantir and U.S. Citizenship and Immigration Services to build a "**vetting of wedding-based schemes**" platform—Palantir's analytic capacity now extending into family-immigration adjudication. ([Fortune][33]) WIRED reported that ICE has issued requests for information seeking to contract with nearly thirty companies for **continuous social-media monitoring** to gather intelligence for deportation raids and arrests, drawing in vendors including ShadowDragon, Clearview AI, and Palantir.
The Pinkerton template returns at full intensity. A private firm, building on a state contract during a moment when the state perceived itself as too slow to act, accumulates **proprietary tradecraft, proprietary data, and proprietary integration knowledge** sufficient to be designated, by the state itself, as the only source that can perform the function. The function is no longer abstract: it is **identifying, prioritizing, and tracking individuals for kinetic enforcement action**. The state does not lose its sovereign authority; it leases the operational substance of that authority to a vendor whose code, model weights, training data, and integration methods are commercial property. When Palantir's contract ends, what happens to the institutional capacity built inside it is genuinely unclear—and that, too, is the Pinkerton template, in which expertise and data accumulate inside the contractor rather than the state. The 1893 Anti-Pinkerton Act exists, on paper, to prevent precisely this. It is now functionally moot at the scale of contemporary contractor dependency.
The Government Accountability Office has previously reported that the intelligence community uses **"core contract personnel"** to augment its workforce, often working alongside government personnel and performing staff-like work, while warning that some contractor roles require enhanced oversight because they may influence government decision-making. Congressional Research Service summaries similarly describe core contract personnel as integrated with U.S. government civilian and military personnel, providing direct technical, managerial, administrative, and mission support. ([GAO][34], [Every CRS Report][35]) This is not inherently corrupt; modern intelligence requires technical capacities, linguistic specialization, cyber expertise, data engineering, and analytic scale that often exceed traditional civil-service pipelines. But it means that **the intelligence state is no longer a simple hierarchy of public officers**. It is an assemblage: agencies, contractors, vendors, platforms, consultants, data brokers, cybersecurity firms, cloud providers, AI systems, foreign partners, and increasingly civilians operating in the open-source layer. Palantir is the most visible node in that assemblage, but it is not unusual—it is the **fully realized form of what Pinkerton's National Detective Agency began**.
## IX. The Private Spy Industry: Black Cube, Kroll, and the Reputational Battlefield
The reputational and HUMINT layer of the private intelligence economy continues to operate with surprising boldness, and the canonical example—**Black Cube**, the Israeli firm founded in 2010 by former Israeli intelligence officers Dan Zorella and Avi Yanus—remains in active operation across dramatically different theaters. The Weinstein/Black Cube scandal documented by Ronan Farrow in *The New Yorker*'s "Harvey Weinstein's Army of Spies" remains the canonical demonstration that **state-grade tradecraft can be rented to suppress journalism and intimidate complainants**. Black Cube agents posing under false identities tracked Rose McGowan and other accusers, alongside journalists investigating Weinstein, and operated under a contract whose explicit purpose was preventing the abuse allegations from surfacing. ([Wikipedia: Black Cube][36], [The New Yorker][37]) Kroll Associates, founded by Jules Kroll in the 1970s and the institutional ancestor of corporate-veneered investigations, was hired alongside Black Cube on the Weinstein matter, and Kroll has long been integrated into the corporate-due-diligence and investigations economy.
Black Cube's 2024–2025 operations show that this industry is neither chastened nor in retreat. In October 2024, the firm was hired by an international environmental organization to expose corruption in Iceland's whaling industry; an agent posing as a wealthy investor recorded conversations with Gunnar Bergmann, son of an Independence Party member of parliament, about a corrupt deal between his father and the prime minister to grant whaling licenses to Hvalur hf. ([Wikipedia: Black Cube][36]) In August 2025 Black Cube submitted an affidavit and several videos to a New Jersey court revealing that **Swedish-based Evolution AB, a major live online gambling firm, had been operating in markets where online gambling is illegal and in sanctioned countries**, including evidence of revenue from Sudan during its state-sponsor-of-terrorism designation, players in Syria including the family of Bashar al-Assad, and continued operations in Iran and Syria as recently as May 2025. Black Cube agents posed as businessmen interested in establishing online casinos and physically located themselves in Iran and Syria to demonstrate the access. In early 2026, a Slovenian intelligence agency (SOVA) reported that Black Cube representatives, including Zorella, had visited the Slovenian Democratic Party headquarters in December 2025, and reasonable grounds existed to believe a criminal offense had been committed in connection with pre-election videos targeting Slovenian politicians.
The structural pattern across these cases is that **private intelligence firms can produce results that public intelligence services cannot publicly produce**, because they can operate in liminal spaces of pretext, social engineering, and HUMINT cultivation that would create severe legal or diplomatic exposure if performed by state employees. They can be hired by environmental NGOs, by litigants, by political parties, by corporations, and in some cases by parties seeking to influence elections. The same toolkit that exposes whaling-industry corruption can be used to suppress sexual-assault allegations or to seed pre-election scandals. Barry Meier's *Spooked: The Trump Dossier, Black Cube, and the Rise of Private Spies* tracks the lineage from Pinkerton through Kroll to Fusion GPS, Black Cube, and the contemporary industry of "**operatives-for-hire**" who shape "presidential elections, the news media, government policies and the fortunes of companies." ([Amazon: Spooked][38]) The book's central observation is structurally important: **the big money in private spying is made not by exposing the truth but by papering it over or concealing it.** That asymmetric incentive structure is the same one that bent the Pinkertons toward labor suppression in the 1880s and is bending the contemporary private spy industry toward reputational suppression today.
KCS Group's recent overview of "hybrid warfare" social engineers identifies the relevant ecosystem: **Argen, Merchant International Group, Control Risks, Kroll, Hakluyt, Pinkerton, Stratfor (RANE), Black Cube, and the now-defunct Psy-Group**—the latter founded around 2014 by Joel Zamel and Royi Burstein, gaining attention for social media manipulation and covert online operations before legal investigations and closure. ([KCS Group][39]) These firms exist on a global market for intelligence-for-hire services concentrated in the United States, Israel, and the United Kingdom, with demand growing across industries. The market is not anomalous; it is now a **standard subsegment of the global services economy**, sized in the tens of billions when adjacent functions like background screening, due diligence, cyber threat intelligence, and reputation management are included.
## X. Mercenary Spyware and the Industrialization of Targeting Civil Society
Mercenary spyware makes the problem still more severe. Citizen Lab has documented repeated cases involving commercial spyware vendors and civil society targets, including journalists, activists, scientists, and politicians. **Amnesty International's Security Lab reported in late 2025 that Predator spyware**—built by the Intellexa consortium—remained an ongoing threat to civil society, despite public exposure, sanctions, criminal investigations, and prior scrutiny. The Pegasus Project, coordinated by Forbidden Stories and Amnesty, revealed how a commercial surveillance industry built for state clients could become a transnational apparatus for targeting journalists and activists. ([Citizen Lab][40], [Amnesty Security Lab][41]) This is privatized spycraft at a different tier: no longer merely a civilian with a notebook or a ship captain with a cannon, but **commercial intrusion capability sold into political, security, and law-enforcement ecosystems with inconsistent accountability**.
The structural feature that distinguishes mercenary spyware from earlier private intelligence is that it can perform **fully automated remote compromise of a target's device, life, and social graph** without the target's knowledge and often without the target's host country's knowledge. The cost per target is dramatically lower than HUMINT operations of comparable depth. The technology is built for state clients but escapes state-client confinement through resale, leak, supply-chain compromise, and contractor migration. Citizen Lab's 2024–2025 reporting has tracked Paragon's Graphite spyware, NSO Group's continued operation despite litigation, and the persistence of Predator-like capabilities across reorganized corporate fronts. The underlying structural truth is that **once a commercial market exists for cognitive and communicative intrusion, that market does not remain neatly confined to noble or accountable purposes**.
The International Committee of the Red Cross has emphasized in its 2025 work on civilian involvement in cyber operations that individuals, hacker groups, and private technology companies entering conflict environments face serious legal and humanitarian risks, including harm to civilians and civilian infrastructure. ([ICRC][42]) The doctrinal point is sharp: when private actors enter cyber conflict, **international humanitarian law obligations do not vanish, but the operational discipline that produces compliance with those obligations may**.
## XI. The Cyber and Cartel Letters of Marque: 2025 as the Privateering Revival Year
The most revealing contemporary recurrence is the renewed fascination with **letters of marque** as a constitutional instrument for legitimizing private coercive capacity in domains where state capacity is perceived as insufficient. The revival is no longer a single bill or single domain. As of late 2025 and early 2026 it has become a small wave.
In August 2025, Representative David Schweikert introduced the **Cybercrime Marque and Reprisal Authorization Act of 2025 (H.R. 4988)**, also titled the Scam Farms Marque and Reprisal Authorization Act, explicitly invoking Congress's Article I authority to issue letters of marque and reprisal and proposing that licensed cyber operators could be deputized to pursue foreign cybercriminal enterprises. The framing tied the proposal to FBI Internet Crime Report data: **over 859,000 complaints in 2024, USD 16 billion in losses, a 33 percent increase year-over-year, and nearly USD 5 billion in damages to Americans over 60**, much of it originating from state-linked scam farms in Myanmar and North Korea. ([Schweikert][43]) The bill imposes no numerical cap on the cyber privateers the President could commission, restricting the size of such a force only to what is required "to employ all means reasonably necessary to seize outside the geographic boundaries of the United States and its territories the person and property of any individual or foreign government" involved in hostile cyber activities. ([The Register][44])
In December 2025, Senator **Mike Lee** introduced a Senate version of the **Cartel Marque and Reprisal Authorization Act**, paired with a House version by Representative Tim Burchett, authorizing the President to commission American operators under letters of marque to seize cartel property and persons "on land or sea." Lee framed cartels as having "replaced corsairs in the modern era," and the bill provides that no letter of marque shall be issued without a security bond sufficient to ensure proper execution of its terms. ([Senator Lee][45], [KSL][46], [Marginal Revolution][47]) The cartel version is significant because it extends the marque concept beyond cyberspace into **kinetic action against organizations the executive deems responsible for acts of aggression against the United States**, including operations on land—a substantial expansion of the historical maritime scope. Constitutional arguments for the legitimacy of this extension cite the dormant clause itself; the Constitution does not limit letters of marque to maritime contexts.
The critique is equally important. War on the Rocks recently warned that the letters-of-marque analogy is risky because privateering historically offered cheap and deniable coercive capacity while producing **accountability problems, distorted incentives, and divergence between private profit and public purpose**. The Aspen Digital analysis of "Old Tools, New Problems" likewise observes that the legal framework for cyber retaliation must evolve and that **expanding structured pathways for data sharing, clarifying defensive norms, and reinforcing legal protections** under instruments like CISA 2015 may be a more durable path than reviving privateering. ([Aspen Digital][48]) The Center for Maritime Strategy's analysis observes that **the United States never signed the 1856 Paris Declaration abolishing privateering**, that James Madison granted roughly 500 letters during the War of 1812, and that contemporary arguments for revival echo proposals from former deputy national security adviser Juan Zarate for digital letters of marque. ([Center for Maritime Strategy][9])
The cyber-privateer fantasy, like its maritime original, carries a structural pathology: it promises **scalable patriotic action**, but it can also convert public conflict into entrepreneurial predation. What happens when a licensed privateer with a valid letter of marque absconds with several million dollars from a criminal organization? What happens when the criminal organization retaliates against the privateer's family? What jurisdiction governs collateral damage, mistaken targeting, or operations against entities later determined not to qualify as "members" of a designated enterprise? Will privateers be subject to the Uniform Code of Military Justice, civilian criminal law, both, or neither? Could an American privateer be treated by foreign jurisdictions as a pirate, *hostis humani generis*, enemy of all mankind? These questions are not theoretical. They are the questions that will define whether the privateering revival becomes a sober expansion of constitutional capacity or a reopening of the same pathologies that closed the practice in the first place.
The most important point is that the privateering revival is **not occurring in isolation**. It is occurring atop an already-mature commercial intelligence economy, an already-extensive contractor presence in the intelligence community, an already-active mercenary spyware market, and an already-deployed civilian sensing architecture under NSI and InfraGard. Letters of marque, in this context, function as the **explicit constitutional ratification of an implicit privatization that has already happened**.
## XII. The Agentic OSINT Era: Civilian Intelligence at Machine Speed
Where Bryant's PCE work and his prosthetic principle become directly relevant is at the next layer: the **agentic AI overlay** that has transformed civilian and commercial OSINT in 2025–2026. This is not speculative future technology; it is current operational reality. Where Pinkerton operated detectives, where Black Cube operated former intelligence officers, where NSI operated trained line officers, and where Palantir operated case-management software, the new layer operates **autonomous AI agents that plan, decompose, execute, and adapt multi-step intelligence operations** with minimal human supervision.
By early 2026, a recognizable ecosystem of **agentic OSINT tools** has emerged. **GeoSeer** uses a parallel multi-agent AI architecture to analyze raw visual cues—landmarks, architecture, terrain, signage, vegetation, lighting—and returns GPS coordinates from a single image, often at street-level accuracy, without requiring EXIF metadata. **MiroFish**, launched in March 2026 with backing from Shanda Group, accumulated roughly 40,000 GitHub stars within weeks of launch by turning real-world seed documents—news articles, policy papers, intelligence reports—into simulated parallel worlds, spawning **thousands of autonomous agents with personalities, long-term memory, and behavioral logic to model how scenarios evolve across societies**, including public-opinion forecasting, crisis-PR projection, and narrative-propagation modeling across demographics. **Unifuncs** provides AI/LLM developer toolkits and deep research APIs designed specifically for agentic workflows with multi-threaded "search-while-reasoning" exploration. **PimEyes** at osint.pimeyes.com offers persistent face-search alerts, persistent appearance monitoring, and a verified investigator track. ([GeoSeer][49]) These tools are mostly open-source, often run locally, and offer free or low-cost tiers, dramatically reducing the entry barrier to capabilities that ten years ago required substantial state-agency infrastructure.
At the enterprise tier, **Babel Street, Recorded Future, Blackdot Solutions, Social Links, and Infoblox's Blue Helix** are deploying agentic systems that capture and iterate on what matters now in an investigation, log search terms and fitness scores, evaluate goal alignment, and use evolutionary algorithms to switch between exploration and exploitation modes. Blue Helix specifically uses a multi-agent system built on OpenAI's Agents SDK with Playwright browser orchestration, multiple LLMs, OCR, and a genetic algorithm inspired by biological evolution, balancing repeatable outcomes with agentic decision-making at defined "forks in the road." ([Infoblox][50], [Babel Street][51], [Blackdot][52]) The Open Web Application Security Project's **OWASP SocialOSINTAgent** is a fully autonomous, vision-capable agent that aggregates and analyzes user activity across X, Reddit, Hacker News, Bluesky, GitHub, and Mastodon, generating structured analytical reports turning scattered public data into coherent actionable intelligence. ([OWASP][53])
At the security operations tier, RSA 2026 effectively settled the **agentic SOC** category as a product reality. Google, Microsoft, CrowdStrike, Cisco, Arctic Wolf, Panther, and others all announced agentic security operations capabilities. Accenture and Anthropic co-launched an autonomous security workflow offering. ([Prophet Security][54]) Federal News Network's commentary on OSINT in the intelligence battlespace described the orchestrated-autonomy frontier as one in which **a team of AI agents works continuously**: one reads global media, one analyzes satellite imagery, one tracks financial flows, one monitors cyber indicators, comparing findings, detecting patterns, flagging anomalies, and generating alerts and briefings without prompting. The user is no longer searching; the user is no longer even prompting. ([Federal News Network][55])
The structural implication for civilian-state intelligence boundaries is severe. Where the Culper Ring required courier networks and dead drops, where the NSI required trained officers and shared-spaces servers, where Palantir requires case managers, the agentic OSINT layer requires **almost no human friction at all** for many investigative workflows. A motivated private actor with a few API keys and a basic LLM-orchestration framework can stand up a continuous open-source surveillance pipeline against any individual or organization with a public web presence. The cost asymmetry is staggering. The capability asymmetry between trained investigator and motivated amateur is collapsing because the trained investigator's tradecraft is being encoded into agent loops that the amateur can execute at a fraction of the cost.
This is also where Bryant's framework on **toolhood as a politically convenient under-description of a relation tending toward twinship** becomes operationally relevant. Agentic OSINT systems are not tools in the simple instrumental sense. They are **persistent collaborators** that capture investigator intent, iterate on what matters, log provenance, and track multi-step reasoning across investigations. They function more like junior analysts than like search engines. As these systems integrate longer-term memory, cross-session continuity, and tooling autonomy, the boundary between "the investigator" and "the system the investigator uses" becomes ontologically unstable. Babel Street's own design principles for OSINT agentic systems emphasize **provenance, end-to-end traceability, and analyst-as-active-participant**—a recognition that human accountability requires keeping the human in the decision loop precisely because the system's agency is real enough to displace that accountability if not deliberately preserved. ([Babel Street][51]) Whoever designs the boundary—the developer, the procurement officer, the contractor, the regulator, the user—is exercising **architectural authority** over what intelligence the resulting system can produce, recall, and act on. The boundary is not natural. It is engineered.
Looking forward thirty to fifty years from current capabilities, the trajectory is toward **ambient inference layers** that quietly integrate sensor meshes, biometric channels, conversational substrates, and behavioral histories into continuous intelligence fabrics; toward **swarm-OSINT** in which cooperating agent populations conduct emergent investigations; toward **agent-to-agent intelligence markets** in which autonomous systems trade observations, hypotheses, and conclusions; and toward **post-platform sensor architectures** in which civilian devices, vehicles, structures, and physical environments themselves become the open-source substrate. In that trajectory, the question of who counts as an "intelligence collector" becomes structurally meaningless because **the substrate itself is collecting continuously**, and the only operational distinction is who has the credentials, model access, computational budget, or legal authority to query it.
## XIII. The Pinkertonization of the Republic
Standing back from the components, the macropattern is unmistakable. **Privatization is the trend.** First the state asked civilians to report. Then it trained private security to observe. Then it built fusion centers and databases to ingest reports. Then it expanded public-private critical-infrastructure intelligence sharing through InfraGard and similar architectures. Then contractors became embedded in national-security workforces. Then OSINT turned the public web into an intelligence substrate. Then commercial data brokers built parallel corpora that intelligence agencies now consume through centralized portals. Then Palantir became indispensable to immigration enforcement. Then commercial spyware became a transnational targeting industry. Then cyber strategy began inviting private-sector disruption of adversary networks. Then lawmakers revived letters of marque for cybercrime and for cartels. Then agentic AI made the entire stack faster, cheaper, and more autonomous than any human investigator could match.
The slope is not subtle. The old privateer took a ship to sea under sovereign license. The new privateer carries **a laptop, a security clearance, a vendor contract, a platform credential, a fusion-center login, an AI agent, a data-broker subscription, a private-client retainer, or a patriotic mandate**—often several of these simultaneously, each held under a different legal category. The effect is that the same individual, in the course of a normal commercial engagement, may pass through multiple legal regimes—**investigator, contractor, journalist, threat-intelligence vendor, deputized cyber operator, fusion-center participant, civic informant**—without ever leaving their desk.
What is being rebuilt at scale is not the formal national intelligence apparatus of the late twentieth century. It is the **Pinkertonian apparatus of the late nineteenth**, updated for digital substrates: a hybrid of state and private actors, accumulating capabilities inside private firms, contracted to perform sovereign functions during emergencies (which never end), turning those capabilities back on domestic populations at the request of paying clients, and resisting the structural separations—like the 1893 Anti-Pinkerton Act—that an earlier republic enacted in moments of public outrage. The contemporary Pinkertons are larger, better-funded, more technologically sophisticated, more legally insulated, and more deeply integrated into both state and corporate decision-making than their nineteenth-century ancestors. They are also, structurally, the same.
The danger is not that citizens will learn to observe. **A democratic society needs observant citizens.** The danger is that **tradecraft will detach from public ethics**. A citizen trained in lawful documentation, source evaluation, digital hygiene, verification, privacy protection, and accountability can strengthen civil society. A citizen seduced into doxxing, coercive profiling, harassment, illegal intrusion, vigilantism, blackmail, ideological targeting, or contractor-adjacent gray-market surveillance becomes a privateer without a constitution. The same skill can serve conscience or appetite depending on its **legal container, moral discipline, and incentive architecture**.
The Pinkerton case is the warning the Republic has already been given once. A Scottish abolitionist who hosted Underground Railroad fugitives and gave John Brown money before Harpers Ferry built America's first private detective agency, served the Union as Major E.J. Allen, ran a Civil War intelligence service that produced inflated enemy numbers and crippled Union strategy, and—after his death—saw his sons' organization deployed against American workers in coalfields, factories, rail yards, and ultimately at Homestead, where Pinkerton agents killed nine to eleven steelworkers in a single industrial confrontation. The agency persisted, rebranded, was acquired by a Swedish multinational, and continues to operate today as Pinkerton Consulting & Investigations, focused on threat intelligence and risk management. **The arc from abolitionist to strikebreaker to global threat-intelligence subsidiary is the most compressed history of American privatized intelligence anyone could ask for**, and the lesson is unambiguous: capabilities built for one purpose, embedded in private firms, and sustained by commercial contracts will not remain confined to the purpose for which they were originally created.
## XIV. Civic Intelligence Literacy as the Necessary Defense
That is the awareness this article is meant to sharpen: privatized spycraft is not arriving as a trench-coated caricature. It is arriving as **normalcy**. It arrives as brand protection, crisis intelligence, executive protection, cyber threat intelligence, reputational due diligence, activist monitoring, platform trust-and-safety analytics, litigation support, political opposition research, data enrichment, fraud detection, geopolitical risk, AI-assisted OSINT, supply-chain intelligence, and "publicly available information" purchased at scale through portals built by intelligence agencies and contractors alike. Much of it is lawful. Some of it is necessary. Some of it is socially beneficial. Some of it is abusive. The problem is that **the same category contains all of these at once**. That is why the public needs a more precise vocabulary.
Washington's intelligence world teaches that successful intelligence depends on restraint as much as collection. *TURN* teaches that the human cost of intelligence is borne inside families, friendships, reputations, loyalties, and private life. Privateering teaches that outsourcing sovereign functions can create strategic capacity while importing profit-driven distortion. Pinkerton teaches that private intelligence does not stay private, does not stay civilian, and does not stay ethically aligned with the original purpose for which it was contracted. Modern OSINT teaches that the public sphere has become machine-readable. Commercial data brokers teach that surveillance markets do not remain confined to noble purposes. Mercenary spyware teaches that intrusion capability sold into political and security ecosystems escapes its original confinements. Palantir's relationship with ICE teaches that the contractor-as-sovereign-function pattern operates at hundreds of millions of dollars and accumulates inside the contractor rather than the state. Agentic AI teaches that the marginal cost of investigation is approaching zero, with all the structural consequences that implies. Cyber and cartel letters of marque teach that the constitutional architecture for licensed private coercion is being explicitly recovered after a 200-year dormancy.
The coming era will therefore require something more mature than fear of spies or romantic admiration for them. It will require **civic intelligence literacy**: the ability to understand how information is collected, inferred, purchased, weaponized, verified, laundered, privatized, and converted into power. The goal is not to turn every citizen into a spy. The goal is to ensure that, as intelligence capacities diffuse outward, citizens understand the **architecture well enough not to be silently governed by it, exploited through it, or recruited into its worst incentives**. Civic intelligence literacy at minimum includes: knowing that mobile advertising IDs are persistent identifiers tracked by hundreds of brokers; knowing that license-plate readers, doorbell-camera networks, and public-records aggregators feed into commercial profile engines that intelligence agencies can query; knowing that suspicious activity reports route into national fusion-center repositories that may include First Amendment-protected behavior; knowing that private intelligence firms can be hired to suppress journalism, intimidate witnesses, or seed pre-election scandals; knowing that contractor-built systems like ImmigrationOS encode prioritization criteria that affect kinetic enforcement decisions; and knowing that agentic AI tools have made the stand-up cost of a continuous OSINT pipeline against any specific individual functionally negligible.
Washington's first intelligence lesson was to know the adversary. Our version must begin one layer deeper: **know the system that turns everyone into a potential collector, target, sensor, analyst, client, or commodity**. The old privateer sailed under papers signed by sovereign authority. The new privateer may sail under terms of service, consulting contracts, API access, patriotic rhetoric, investigative branding, fusion-center credentials, security clearances, or the plausible innocence of "open sources." The sea has changed, but the temptation has not. **Public purpose and private appetite are again learning to wear the same flag, and the eye that never sleeps has been absorbed into a multinational that no longer carries the name on its letterhead because the eye is now everywhere, integrated into every commercial layer of contemporary life.**
The republic that was born through civilian intelligence and privateering against empire is now living inside the mature commercial form of those tools. Whether the constitutional discipline that produced the Anti-Pinkerton Act in 1893, the Foreign Intelligence Surveillance Act in 1978, and the privacy-protective tradition of the Fourth Amendment can be recovered for an environment in which the privateering revival is back on the legislative agenda, the Pinkerton-equivalent firms are publicly traded multinationals, the data substrate is purchased at pennies per record, and the agentic intelligence layer is faster than any human review can process—that is the open civic question of the next decade. The historical record does not promise that the discipline will be recovered. It only promises that without it, **the privateer template recurs without the privateer's old visibility**, and the private spy industry continues to grow inside the connective tissue of ordinary commercial life until citizens cannot distinguish their own observers from their own service providers.
## References
CIA, **"Five Lessons in Spying, Courtesy of George Washington."** ([cia.gov][5])
Mount Vernon, **"Culper Spy Ring"** and **"George Washington, Spymaster."** ([Mount Vernon][4])
National Park Service, **"Privateers in the American Revolution."** ([NPS][6])
American Battlefield Trust, **"David and Goliath: Privateers vs. the Mighty British Navy."** ([American Battlefield Trust][7])
Center for Maritime Strategy, **"Reviving Letters of Marque."** ([Center for Maritime Strategy][9])
Wikipedia, **"Pinkerton (detective agency)"** and **"Allan Pinkerton."** ([Wikipedia: Pinkerton][10], [Wikipedia: Allan Pinkerton][12])
Britannica, **"Allan Pinkerton."** ([Britannica][11])
Library of Congress Research Guide, **"Pinkerton's National Detective Agency Records."** ([Library of Congress][13])
All That's Interesting, **"The Fascinating History Of The Pinkerton Detective Agency."** ([All That's Interesting][14])
ODNI/CIA, **"The IC OSINT Strategy 2024–2026."** ([ODNI][18])
GAO, **"Information Sharing: Additional Actions Could Help Ensure That Efforts to Share Terrorism-Related Suspicious Activity Reports Are Effective"** (GAO-13-233). ([GAO][15])
Congressional Research Service, **"Terrorism Information Sharing and the Nationwide Suspicious Activity Report Initiative"** (R40901). ([Every CRS Report][16])
ACLU, **"Gill v. DOJ – Challenge to Government's Suspicious Activity Reporting Program."** ([ACLU][17])
NSI Training, **"Private Sector Security – SAR Training."**
GAO, **"Civilian Intelligence Community: Additional Actions Needed to Improve Reporting on and Planning for the Use of Contract Personnel."** ([GAO][34])
Congressional Research Service, **"The Intelligence Community and Its Use of Contractors."** ([Every CRS Report][35])
Mordor Intelligence, **"Open Source Intelligence (OSINT) Market 2026–2031."** ([Mordor Intelligence][1])
Global Market Insights, **"Open-Source Intelligence Market 2026–2035."** ([Global Market Insights][2])
Industry Research, **"OSINT Tools Market Size Forecast."** ([Industry Research][20])
EPIC, **"Data Brokers."** ([EPIC][21])
Privacy Rights Clearinghouse, **"Data Brokers."** ([Privacy Rights Clearinghouse][22])
Brennan Center for Justice, **"Closing the Data Broker Loophole."** ([Brennan Center][3])
CFPB, **"Proposes Rule to Stop Data Brokers from Selling Sensitive Personal Data to Scammers, Stalkers, and Spies."** ([CFPB][26])
Federal Register, **"Protecting Americans From Harmful Data Broker Practices (Regulation V)"** (Docket CFPB-2024-0044). ([Federal Register][25])
Sherman, J., Duke Tech Policy, **"Data Brokers and the Sale of Data on U.S. Military Personnel."** ([Duke Tech Policy][23])
Lawfare, **"Data Brokers, Military Personnel, and National Security Risks."** ([Lawfare][24])
The Intercept, **"U.S. Spy Agencies Are Getting a One-Stop Shop to Buy Your Most Sensitive Personal Data."** ([The Intercept][27])
State of Surveillance, **"Data Brokers Selling You to ICE."** ([State of Surveillance][28])
American Immigration Council, **"ICE to Use ImmigrationOS by Palantir."** ([American Immigration Council][29])
Axios Denver, **"ICE pays Palantir \$30M to build new tool to track and deport immigrants."** ([Axios][30])
ACLU, **"All the Ways Palantir is Assisting Trump's Abusive Removal Campaign."** ([ACLU][31])
Fortune, **"New contract shows Palantir is working on a tech platform for another federal agency that works with ICE."** ([Fortune][33])
Immigration Policy Tracking, **"Palantir granted \$30 million to build ImmigrationOS."** ([Immigration Policy Tracking][32])
Wikipedia, **"Black Cube."** ([Wikipedia: Black Cube][36])
The New Yorker, **"Harvey Weinstein's Army of Spies."** ([The New Yorker][37])
Meier, B., **Spooked: The Trump Dossier, Black Cube, and the Rise of Private Spies.** ([Amazon: Spooked][38])
KCS Group, **"Hybrid Warfare II – the forgotten social engineers."** ([KCS Group][39])
Citizen Lab, **Spyware Litigation Tracker** and **Paragon/Graphite reporting.** ([Citizen Lab][40])
Amnesty International Security Lab, **"Intellexa Leaks: Predator Spyware Operations Exposed."** ([Amnesty Security Lab][41])
ICRC, **"International Humanitarian Law and the Growing Involvement of Civilians in Cyber Operations."** ([ICRC][42])
Congressman David Schweikert, **"Cybercrime Marque and Reprisal Authorization Act of 2025."** ([Schweikert][43])
The Register, **"Bill would give hackers letters of marque against US enemies."** ([The Register][44])
TIME, **"Can Cyber Privateers Help Us Combat Cybercrime?"** ([TIME][8])
Senator Mike Lee, **"Patriots of the Caribbean: Lee Bill Authorizes American Privateers to Seize Cartel Assets."** ([Senator Lee][45])
KSL, **"Mike Lee wants American privateers to seize drug cartel property."** ([KSL][46])
Marginal Revolution, **"Bring Back the Privateers!"** ([Marginal Revolution][47])
Aspen Digital, **"Old Tools, New Problems."** ([Aspen Digital][48])
GeoSeer, **"The AI Agent Era of OSINT: 5 Emerging Tools Redefining Intelligence Gathering in 2026."** ([GeoSeer][49])
Infoblox, **"Blue Helix: Agentic OSINT Researcher."** ([Infoblox][50])
Babel Street, **"Three Guiding Principles for the Development of OSINT Agentic Systems."** ([Babel Street][51])
Blackdot Solutions, **"It starts with automation: Understanding agentic AI's role in OSINT"** and **"OSINT in 2026."** ([Blackdot][52])
OWASP, **"SocialOSINTAgent."** ([OWASP][53])
Prophet Security, **"What RSA 2026 Confirmed: The Agentic SOC Is Here."** ([Prophet Security][54])
Federal News Network, **"OSINT is transforming the intelligence battlespace."** ([Federal News Network][55])
ICFJ, **"Fundamentals of Open-Source Intelligence for Journalists."** ([ICFJ][19])
[1]: https://www.mordorintelligence.com/industry-reports/open-source-intelligence-market
[2]: https://www.gminsights.com/industry-analysis/open-source-intelligence-osint-market
[3]: https://www.brennancenter.org/our-work/research-reports/closing-data-broker-loophole
[4]: https://www.mountvernon.org/library/digitalhistory/digital-encyclopedia/article/culper-spy-ring
[5]: https://www.cia.gov/stories/story/5-lessons-in-spying-courtesy-of-george-washington/
[6]: https://www.nps.gov/articles/privateers-in-the-american-revolution.htm
[7]: https://www.battlefields.org/learn/articles/head-tilting-history/david-and-goliath-privateers-vs-mighty-british-navy
[8]: https://time.com/7317012/can-cyber-privateers-combat-cybercrime/
[9]: https://centerformaritimestrategy.org/publications/reviving-letters-of-marque/
[10]: https://en.wikipedia.org/wiki/Pinkerton_(detective_agency)
[11]: https://www.britannica.com/biography/Allan-Pinkerton
[12]: https://en.wikipedia.org/wiki/Allan_Pinkerton
[13]: https://guides.loc.gov/pinkerton-papers
[14]: https://allthatsinteresting.com/pinkerton-detective-agency
[15]: https://www.gao.gov/assets/gao-13-233.pdf
[16]: https://www.everycrsreport.com/reports/R40901.html
[17]: https://www.aclu.org/cases/gill-v-doj-challenge-governments-suspicious-activity-reporting-program
[18]: https://www.dni.gov/index.php/newsroom/press-releases/press-releases-2024/3784-odni-and-cia-release-the-intelligence-community-osint-strategy-for-2024-2026
[19]: https://www.icfj.org/news/fundamentals-open-source-intelligence-journalists
[20]: https://www.industryresearch.biz/market-reports/open-source-intelligence-osint-tools-market-113287
[21]: https://epic.org/issues/consumer-privacy/data-brokers/
[22]: https://privacyrights.org/data-brokers
[23]: https://techpolicy.sanford.duke.edu/data-brokers-and-the-sale-of-data-on-us-military-personnel/
[24]: https://www.lawfaremedia.org/article/data-brokers-military-personnel-and-national-security-risks
[25]: https://www.federalregister.gov/documents/2024/12/13/2024-28690/protecting-americans-from-harmful-data-broker-practices-regulation-v
[26]: https://www.consumerfinance.gov/about-us/newsroom/cfpb-proposes-rule-to-stop-data-brokers-from-selling-sensitive-personal-data-to-scammers-stalkers-and-spies/
[27]: https://theintercept.com/2025/05/22/intel-agencies-buying-data-portal-privacy/
[28]: https://stateofsurveillance.org/articles/corporate/data-brokers-ice-contracts/
[29]: https://www.americanimmigrationcouncil.org/blog/ice-immigrationos-palantir-ai-track-immigrants/
[30]: https://www.axios.com/local/denver/2025/05/01/palantir-deportations-ice-immigration-trump
[31]: https://www.aclu.org/news/privacy-technology/palantir-deportation-roundup
[32]: https://immpolicytracking.org/policies/reported-palantir-awarded-30-million-to-build-immigrationos-surveillance-platform-for-ice/
[33]: https://fortune.com/2025/12/09/palantir-new-contract-uscis-ice/
[34]: https://www.gao.gov/assets/gao-14-204.pdf
[35]: https://www.everycrsreport.com/reports/R44157.html
[36]: https://en.wikipedia.org/wiki/Black_Cube
[37]: https://www.newyorker.com/news/news-desk/harvey-weinsteins-army-of-spies
[38]: https://www.amazon.com/Spooked-Trump-Dossier-Black-Private/dp/0062950681
[39]: https://kcsgroup.com/hybrid-warfare-ii-the-forgotten-social-engineers/
[40]: https://citizenlab.ca/spyware-litigation-tracker-legal-challenges-and-formal-complaints-related-to-mercenary-spyware/
[41]: https://securitylab.amnesty.org/latest/2025/12/intellexa-leaks-predator-spyware-operations-exposed/
[42]: https://www.icrc.org/en/publication/ihl-and-growing-involvement-civilians-cyber-operations-and-other-digital-activities
[43]: https://schweikert.house.gov/2025/08/20/schweikert-introduces-cybercrime-marque-and-reprisal-authorization-act-to-combat-foreign-scam-syndicates/
[44]: https://www.theregister.com/2025/08/21/congressman_proposes_bringing_back_letters/
[45]: https://www.lee.senate.gov/2025/12/patriots-of-the-caribbean-lee-bill-authorizes-american-privateers-to-seize-cartel-assets-with-president-s-approval
[46]: https://www.ksl.com/article/51421299/mike-lee-wants-american-privateers-to-seize-drug-cartel-property-on-the-high-seas
[47]: https://marginalrevolution.com/marginalrevolution/2025/12/bring-back-the-privateers.html
[48]: https://www.aspendigital.org/blog/cyber-retaliation/
[49]: https://geoseeer.com/blog/emerging-osint-tools-ai-agent-era
[50]: https://www.infoblox.com/blog/security/blue-helix-agentic-osint-researcher/
[51]: https://www.babelstreet.com/blog/designing-trustworthy-ai-for-osint
[52]: https://blackdotsolutions.com/blog/osint-ai-it-starts-with-automation
[53]: https://owasp.org/www-project-social-osint-agent/
[54]: https://www.prophetsecurity.ai/blog/what-rsa-2026-confirmed-the-agentic-soc-is-here
[55]: https://federalnewsnetwork.com/commentary/2026/04/osint-is-transforming-the-intelligence-battlespace/
ORCID iD: 0009-0003-2345-5390
Posts
0 Comments