Kessler's Gambit
Martin Kessler began writing the white paper on February 14, twelve days after his statement appeared in The Washington Post and The New York Times. He wrote in his study in Bethesda, Maryland, in a room that contained a desk, a chair, a laptop, and eleven linear feet of case reporters on a single shelf. No photographs. No diplomas. He had graduated first in his class at Harvard Law School, and the diploma was in a box in his basement, still in the mailing tube.
He started with the title. Titles mattered in legal scholarship. A title that signaled a polemic would be dismissed by the academy. A title that signaled empirical work would attract the right readers. He typed:
LEGAL ARBITRAGE AT SCALE: A STRUCTURAL ANALYSIS OF PERMISSIBLE ECONOMIC EXTRACTION UNDER UNITED STATES FEDERAL AND STATE LAW
Forty-seven minutes on the title. Twenty-three minutes on the abstract. Six weeks on the body.
The abstract read:
This paper documents six categories of economic extraction, collectively generating approximately $69.3 billion in annual economic damage, conducted entirely within the boundaries of existing federal and state law. For each category, the paper identifies the specific statutory provisions and judicial precedents that permit the activity, the structural characteristics of the legal framework that make prohibition difficult, and proposed legislative remedies. The paper argues that the gap between permissible conduct and just outcomes is not an anomaly but a structural feature of rules-based legal systems, and that the most efficient path to reform is full disclosure of the mechanisms by which that gap is exploited.
He did not use the word "Consortium." He did not name any entity. He did not name himself. The paper was written in the third person passive, the voice of scholarship, as though the author were merely an observer describing a system that existed independently of anyone's design.
This was technically accurate. The legal system existed independently of his design. He had merely read it.
Rachel Tan reviewed the draft on March 2. She sat across from Kessler in his study, reading on her tablet, making no marks.
"Section 2.3," she said. "You cite the specific regression coefficients for LEGAL_PROB. Those are from the algorithm."
"Those are from the ProPublica article. They published the coefficients. Public domain."
"The recalibration methodology in Section 2.4 is not public. Marcus leaked the document, and ProPublica published selected portions. You are publishing the full technical specification."
"I wrote the specification. It is my intellectual property."
Tan set the tablet down. "Martin. The six operations collectively employ 46,000 people. Publishing the complete methodology allows anyone to replicate the system. You are describing, in precise detail, how to build a machine that causes $69.3 billion in annual damage."
"I am describing, in precise detail, how the legal system permits $69.3 billion in annual damage. And I am providing specific legislative remedies for each mechanism. Section by section. If Congress reads this paper and implements the remedies, the machine stops."
"Congress will not read this paper."
"Then the damage is not my responsibility. The damage is the law's responsibility. I am simply providing the documentation."
Tan picked up the tablet again. She read for another forty minutes. When she finished, she asked one question.
"Have you consulted the operation leads?"
"No."
"They will view this as a betrayal."
"They will view this as a risk to their revenue. Those are different things. The operations are legal. They will remain legal until the law changes. Publishing the methodology does not change the law. It provides a roadmap for changing the law. If the roadmap is followed, the operations will need to adapt. If it is not followed, nothing changes."
"And if the roadmap is followed but incompletely? If Congress patches three of the six mechanisms and leaves three open?"
"Then the system adapts to operate through the three remaining mechanisms. Which is what it would do regardless. The system adapts to legal change. That is its design."
Tan stood up. "I'll have the operation leads notified as a courtesy. Twenty-four hours before publication."
"Seventy-two."
She nodded and left. Kessler returned to Section 4.
He uploaded the paper to the Social Science Research Network on March 19 at 6:00 AM Eastern. SSRN was the standard repository for legal scholarship. Papers were freely downloadable. Authors received no payment. The platform assigned a DOI and an abstract ID. Kessler's paper received abstract ID 4897312. He selected the Legal Studies category and the subcategories of Regulatory Law, Law and Economics, and Public Policy. He attached a Creative Commons Attribution 4.0 International license, which permitted anyone to copy, redistribute, adapt, and build upon the work for any purpose, including commercial use.
He set the author field to: Martin D. Kessler, J.D., Kessler & Associates, LLP.
No co-authors. No acknowledgments. No funding disclosure.
He clicked Publish.
Then he made coffee, poured it into a ceramic mug with no markings, and sat in his study reading Guido Calabresi's "The Costs of Accidents." He had first read Calabresi as a 1L at Harvard in 1993. The argument, that the law should minimize the sum of accident costs and accident prevention costs, had struck him then as obvious. It struck him now as incomplete. Calabresi had assumed that the law sought to minimize harm. Kessler's thirty-year career had demonstrated that the law did not seek anything. The law was a set of rules. Rules did not seek. Rules permitted and prohibited. What happened in the space between permission and prohibition was not the law's concern. It was everyone else's.
Elena learned about the white paper from James Okafor. He called at 7:41 AM on March 19, before she had finished her first cup of coffee.
"He published it."
"Published what?"
"The whole thing. The complete architecture. SSRN. Freely downloadable. Creative Commons."
Elena set her coffee down. She opened her laptop, typed "Kessler Legal Arbitrage SSRN" into Google, and found the paper in four seconds. She clicked Download. The PDF loaded. One hundred and forty-seven pages.
She read the abstract. Read it again.
"James."
"I know."
"He's publishing the blueprint."
"He's publishing everything. Every operation. Every statute. Every case citation. Every mechanism. Every loophole. And the remedies. Section by section. Legislative language. Specific bill text. He wrote the reform legislation and published it alongside the exploitation methodology."
Elena scrolled through the table of contents. Six sections, one per operation. Each section followed the same structure:
1. Operative Mechanism (what the operation does) 2. Statutory Basis (the specific laws that permit it) 3. Judicial Precedent (the case law that reinforces it) 4. Structural Barriers to Prohibition (why current law cannot address it) 5. Proposed Legislative Remedy (specific statutory language to close the gap) 6. Predicted Adaptation Pathways (how the operation would evolve in response to the remedy)
She stopped at item six. She read it again. Predicted Adaptation Pathways.
He was not just publishing the exploit. He was publishing the exploit, the fix, and the next exploit that would emerge after the fix was applied.
"James, Section 6 of each chapter. He's predicting how the system adapts."
"I saw it. It's the most honest thing I've ever read and the most dangerous thing I've ever read, and I cannot tell whether those are the same thing."
Elena started reading Section 1: MINOTAUR. Patent Assertion Entities and the Strict Liability Framework of 35 U.S.C. Section 271.
The section opened with a literature review. Kessler cited Colleen Chien's 2012 study of patent assertion entities in the Santa Clara Law Review, James Bessen and Michael Meurer's "Patent Failure" from MIT Press, and the 2013 White House report on patent assertion entities. All real. All publicly available. All supporting the same conclusion: patent trolling was a known problem, documented extensively in academic literature, and Congress had failed to address it despite multiple legislative proposals, including the Innovation Act of 2013 (H.R. 3309, 113th Congress) and the PATENT Act of 2015 (S. 1137, 114th Congress), both of which passed one chamber and died in the other.
The operative mechanism was described with the precision of engineering documentation. Step one: acquire patents with broad claims from financially distressed companies or individual inventors, paying between $50,000 and $500,000 per patent depending on claim breadth and remaining term. Step two: identify potential infringement targets through automated scanning of product documentation, marketing materials, and source code repositories. Step three: send demand letters through single-purpose LLCs incorporated in the Eastern District of Texas (prior to TC Heartland LLC v. Kraft Foods Group Brands LLC, 581 U.S. 258 (2017)) or the District of Delaware (after TC Heartland restricted venue to states of incorporation). Step four: set demand amounts below the expected cost of litigation, typically between $50,000 and $300,000 for small to mid-size companies. Step five: collect settlements.
Each step cited the specific statutory provision that authorized it. Patent acquisition: legal under 35 U.S.C. Section 261, which provides that patents are assignable. Demand letters: legal under the First Amendment's Petition Clause, as affirmed in Professional Real Estate Investors, Inc. v. Columbia Pictures Industries, Inc., 508 U.S. 49 (1993). Venue selection: governed by 28 U.S.C. Section 1400(b) and the TC Heartland interpretation. Settlement demands: a private negotiation between parties, not subject to regulatory oversight unless accompanied by fraud.
Elena read the section in forty minutes. Every citation was correct. She verified three at random using Westlaw. All three were real cases with correct reporter citations and correct holdings.
She moved to Section 3: GOLEM.
By noon on March 19, the paper had been downloaded 4,200 times. By midnight, 31,000. By the end of the first week, 147,000.
The first academic response came from Professor Lawrence Chen at Yale Law School. Chen, who held the Sterling Chair in Law and Political Economy, published a two-page response on his faculty blog at 3:15 PM on the day of publication. He called the paper "the most significant contribution to the literature on regulatory arbitrage since Victor Fleischer's 2009 article in the Texas Law Review." He noted that the paper's structure, pairing exploitation mechanisms with specific remedies, was "unprecedented in legal scholarship" and that the inclusion of predicted adaptation pathways was "either an act of extraordinary intellectual honesty or an act of extraordinary arrogance, and I am not certain the distinction matters."
The second response came from Professor Margaret Walsh at Stanford Law School, published in the Stanford Law Review Online on March 22. Walsh's response was titled "The Penetration Tester's Dilemma." She drew an analogy to the cybersecurity practice of "full disclosure," in which security researchers publish the technical details of software vulnerabilities to pressure vendors into patching them. The analogy was precise: in cybersecurity, the full disclosure debate had raged since the 1990s, with advocates arguing that publishing vulnerabilities forced faster fixes and critics arguing that publication enabled exploitation by malicious actors before patches could be deployed. The Computer Fraud and Abuse Act, 18 U.S.C. Section 1030, had never been used to prosecute a security researcher for publishing vulnerability research, and the Department of Justice had issued formal guidance in 2022 stating that good-faith security research should not be prosecuted under the CFAA.
Walsh argued that Kessler's white paper occupied an analogous position: the publication of legal vulnerabilities to pressure the legislative vendor (Congress) into patching the legal code. "The difference," Walsh wrote, "is that software vulnerabilities can be patched in days. Legislative vulnerabilities take years. And during those years, anyone who downloads the paper has the complete instructions for exploitation."
The third response, also on March 22, came from Professor David Bernstein at George Mason University's Antonin Scalia Law School. Bernstein's response was one paragraph: "Kessler has described legal activity. Publishing a description of legal activity is also legal activity. The fact that people find the described activity offensive does not create a basis for prohibiting the description. See Brandenburg v. Ohio, 395 U.S. 444 (1969), in which the Supreme Court held that advocacy of illegal conduct is protected by the First Amendment unless it is directed to inciting imminent lawless action and is likely to produce such action. Kessler is not advocating illegal conduct. He is describing legal conduct. This is the most protected category of speech that exists."
Elena read all three responses at her desk in Vienna, Virginia. She noted that none of them discussed the actual people affected by the operations. The academic debate was about disclosure norms, about the ethics of publishing, about First Amendment doctrine and cybersecurity analogies. The debate was about Kessler. It was not about the 46,000 agents, the $69.3 billion, or the woman in Akron who was dead.
She opened her own document. The one titled WHAT THE LAW CANNOT DO. She started a new section: WHAT SCHOLARSHIP CANNOT SEE.
Marcus Cole learned about the white paper from Patricia Huang. She called him on March 20, the day after publication, and asked him to come to her office.
He sat across her desk. She had the paper open on her screen, scrolled to Section 3.4.
"This is your section," she said.
Section 3.4 was titled: "Optimization of Collection Enforcement Within Fair Debt Collection Practices Act Safe Harbors." It was fourteen pages. It described HYDRA's operations with the clinical precision of a systems architecture document. The input variables were listed in a table: INC_GROSS, EXP_HOUSING, LEGAL_PROB, HEALTH_FLAG, DEP_COUNT. The optimization function was described mathematically. The objective: maximize portfolio recovery rate subject to the constraint that all collection actions remain within the boundaries of 15 U.S.C. Sections 1692 through 1692p.
The section cited the FDCPA's specific safe harbors. Section 1692c(a)(1), which permits contact at the debtor's usual place of abode. Section 1692e, which prohibits false, deceptive, or misleading representations, but whose definition of "misleading" has been narrowed by courts including Jerman v. Carlisle, McNellie, Rini, Kramer & Ulrich LPA, 559 U.S. 573 (2010), which held that a bona fide error defense is not available for legal errors. Section 1692f, which prohibits unfair practices, but whose courts have interpreted "unfair" narrowly, requiring a showing of conduct that is "unconscionable" rather than merely aggressive.
Marcus read the section. He recognized the variable names. He recognized the coefficients. He recognized the targeting methodology. He had implemented this system for seven years. He had been employee number 4,312 of 46,000. And now the system he had implemented was described in a white paper with a Creative Commons license, downloadable by anyone, reproducible by anyone.
Footnote 247 appeared on page 89. It read: "The phenomenon of convergent targeting, in which multiple operations simultaneously optimize against the same household, represents the system's most efficient extraction mode and its most significant ethical exposure. A single household may simultaneously face patent assertion litigation (MINOTAUR), wage garnishment (HYDRA), and algorithmically determined rent increases (CHIMERA), each independently optimized and each independently legal. The aggregate effect is catastrophic, but no single actor bears responsibility for the aggregate because no single actor controls the aggregate. See generally the problem of cumulative harm in tort law, discussed in Wright, Causation in Tort Law, 73 California Law Review 1735 (1985)."
The footnote did not mention Carla Simmons by name. It did not need to. Marcus knew who the footnote described. He knew the address. He knew the three operations. He knew the aggregate effect.
"He wrote a footnote," Marcus said.
"I'm sorry?"
"She's a footnote. She died, and she's a footnote in an academic paper."
Huang said nothing. She had learned, in sixteen years of legal aid work, that some statements did not require a response.
Marcus read the footnote again. "Convergent targeting." The phrase reduced a woman's death to a technical category. It was precise. It was accurate. It was, in the way that Kessler was always accurate, completely truthful. Convergent targeting was exactly what had happened. Three optimization functions had converged on the same household, not through malice but through mathematics, and the result was that a woman in Akron, Ohio, had decided that the optimal solution to the system of equations that constituted her life was to remove herself as a variable.
Kessler would not have described it that way. Kessler would have described it the way the footnote described it: as a phenomenon. As an ethical exposure. As a topic for further research.
Marcus closed the paper. Huang waited.
"What does this mean for my cases?" he asked.
"Possibly something significant. The white paper publishes the complete methodology of the system you disclosed. Kessler's own description of the system is more detailed than anything you provided. If everything Marcus Cole disclosed is now publicly available through the architect's own publication, the trade secret claims become substantially weaker. A trade secret that has been voluntarily disclosed by its owner is no longer a trade secret. Under the Uniform Trade Secrets Act and the federal Defend Trade Secrets Act, trade secret protection requires that the owner take reasonable measures to maintain secrecy. Publishing on SSRN with a Creative Commons license is the opposite of maintaining secrecy."
"So the lawsuits go away?"
"The Ohio case becomes much stronger for you. I can file a motion to dismiss based on the voluntary disclosure doctrine. Delaware is less certain, because the Delaware Chancery has its own standards for trade secret analysis. New York, I don't know. But the white paper undercuts the foundation of all three claims."
Marcus sat with this for approximately thirty seconds. He processed it the way he processed most information: not through analysis but through the physical sensation of a weight shifting. The weight did not disappear. It redistributed.
"He didn't do this for me," Marcus said.
"No," Huang said. "He did not."
Kessler received his first replication inquiry on April 3, fifteen days after publication. It arrived as a letter, hand-delivered by courier to the offices of Kessler and Associates in Bethesda. The letter was on the stationery of a law firm in London: Carrington Hyde LLP, 15 Old Broad Street, EC2N 1DW. The letter was three paragraphs.
The first paragraph expressed admiration for the scholarship. The second paragraph noted that the regulatory frameworks described in the white paper had analogues in the United Kingdom and the European Union, specifically in the Financial Services and Markets Act 2000, the UK Consumer Credit Act 1974, and the EU Consumer Rights Directive 2011/83/EU. The third paragraph asked whether Dr. Kessler would consider a consultancy engagement to advise on the adaptation of the documented methodology to the European regulatory environment.
Kessler read the letter at his desk. He read it once for content and once for implication. The content was a business inquiry. The implication was that a London firm had read a freely available academic paper, identified the economic opportunity described within it, and decided to build its own version.
This was the predicted outcome. Section 7 of the white paper, titled "Replication Risk and the Case for Preemptive Reform," had described this scenario explicitly. Kessler had written:
The methodology documented in this paper is not proprietary. The legal vulnerabilities it identifies are structural features of every common-law jurisdiction that separates legislative and judicial functions. Any jurisdiction with (a) a rules-based legal system, (b) independent regulatory agencies, (c) private litigation rights, and (d) weak coordination between regulatory domains is susceptible to the same class of exploitation. The publication of this paper will accelerate replication. This is an intended consequence. The author's position is that replication at scale is the most efficient mechanism for generating the political will necessary for structural reform. A single system exploiting legal gaps in a single country can be characterized as an anomaly. Multiple systems exploiting identical gaps across multiple jurisdictions constitute evidence of a structural defect.
Elena would have called this rationalization. Kessler did not distinguish between rationalization and analysis. Both were frameworks applied to facts. Both produced conclusions. Both could be correct.
He placed the letter in a drawer. He did not respond. He did not need to. The paper was under a Creative Commons license. Anyone could build on it. That was the point.
On April 7, Senator Katherine Hartwell, Democrat of Connecticut and chair of the Senate Committee on Banking, Housing, and Urban Affairs, introduced S. 2847, the Economic Exploitation Prevention Act. The bill's text ran to 312 pages. Its provisions tracked, section by section, the legislative remedies proposed in Kessler's white paper.
Section 101 amended 35 U.S.C. Section 271 to create a rebuttable presumption of bad faith for patent assertion entities that had no business operations other than patent licensing, effective only for entities filing more than ten patent infringement claims per year. Section 201 imposed a financial transaction tax of 0.1 percent on all securities transactions executed by high-frequency trading firms, defined as firms executing more than 10,000 trades per day, modeled on the European Commission's proposed Financial Transaction Tax directive from 2013. Section 301 created a federal anti-SLAPP statute, codifying the standard from the California Code of Civil Procedure Section 425.16 and providing for mandatory fee-shifting against plaintiffs whose claims are found to arise from the defendant's exercise of free speech rights. Section 401 amended the Lobbying Disclosure Act, 2 U.S.C. Sections 1601 through 1614, to require disclosure of all lobbying expenditures exceeding $5,000, a reduction from the existing $13,000 quarterly threshold.
Section 501 amended the Fair Debt Collection Practices Act to cap wage garnishment at 10 percent of disposable earnings, down from the 25 percent permitted under 15 U.S.C. Section 1673(a). Section 601 established a federal cap on institutional ownership of single-family residential properties, limiting any single entity and its affiliates to ownership of no more than 1,000 units within any single metropolitan statistical area, as defined by the Office of Management and Budget.
Kessler read the bill text on the morning it was introduced. He read it the way he read all legislation: as code. He identified seven gaps in the first reading.
Gap one: Section 101's threshold of ten patent infringement claims per year could be circumvented by distributing claims across subsidiary entities, each filing nine. Gap two: Section 201's definition of "high-frequency trading firm" was based on trade count, not strategy, meaning firms could reduce trade frequency while increasing trade size to maintain equivalent extraction. Gap three: Section 301's anti-SLAPP standard adopted California's framework but did not address the six-month discovery window that California courts had recognized under Baral v. Schnitt, 1 Cal.5th 376 (2016), during which plaintiffs could still impose significant litigation costs. Gap four: Section 401's reduced lobbying disclosure threshold was still measured quarterly, meaning expenditures could be structured in weekly increments below the per-report threshold.
He stopped at four. The remaining three were in Sections 501 and 601. He would document them later, in a supplemental paper, which he would publish on SSRN after the bill advanced through committee.
He was not surprised by the gaps. He had predicted them. Section 6 of each chapter in the white paper had identified the adaptation pathways that would remain available after the proposed remedies were enacted. The bill's drafters, presumably Senator Hartwell's staff attorneys and the Legislative Counsel's office, had read Section 5 of each chapter and translated the remedies into statutory language. They had not read Section 6.
Nobody ever read Section 6.
Elena read the bill text the same morning. She recognized the structure immediately. Kessler's remedies, translated into legislative language, arranged in the standard format of an omnibus bill. She verified three provisions against the white paper. Exact matches. The bill was the white paper's Section 5, formatted as law.
She looked for Section 6. For the adaptation pathways. For the predicted responses.
They were not in the bill. Bills do not include predictions of their own failure.
Elena opened the white paper. She reread Section 6 of each chapter. She made a list. Seven adaptation pathways. Seven gaps. Seven places where the machine would continue to operate after the bill was enacted, if it was enacted, if it survived committee, if it survived floor debate, if it survived amendment, if it survived the conference committee, if the President signed it, and if the courts upheld it.
She counted the conditional steps between the bill's introduction and its effective implementation. She counted nine. Each step provided an opportunity for BASILISK to intervene. And even if all nine steps were completed successfully, seven gaps would remain.
She thought about what Kessler had said in his February statement. The gap between legal and just is permanent. It is inherent in the structure of any system that attempts to regulate human behavior through written rules.
She had dismissed it as a deflection. She was beginning to suspect it was a description.
Marcus did not read the bill. He read about it on his phone, in a CNN article, while eating a sandwich at the kitchen table in his apartment in Reynoldsburg. The article quoted Senator Hartwell, three legal scholars, and a spokesperson for the National Association of Manufacturers. It did not quote Kessler. It did not quote Marcus. It mentioned Marcus in the eleventh paragraph, identifying him as "the former debt collector whose leaked documents sparked the Department of Justice investigation." The article did not mention Patricia Huang. It did not mention the three lawsuits. It did not mention that Marcus's savings had dropped from $47,000 to $11,400 in legal fees.
He set the phone down. Picked it up. Set it down again.
The bill would take months to pass. Years, maybe. His lawsuits would be resolved before the bill became law. His money would run out before the bill became law. The bill was about the system. The system was about everyone. Marcus was about Marcus, and Marcus had $11,400 and three lawsuits and a one-bedroom apartment and a phone with a CNN article about a bill inspired by a white paper published by the man who had designed the machine that had destroyed a woman's life that Marcus had helped implement and then disclosed and was now being sued for disclosing.
He finished the sandwich. He washed the plate. He sat on the couch and looked at the wall.
The wall had nothing on it. He had not hung anything since moving to Reynoldsburg. The apartment was temporary. Everything about his life since leaving Meridian was temporary. He was waiting for the lawsuits to end, waiting for the bill to pass, waiting for something to change. He was beginning to understand that waiting was not a strategy. It was the absence of a strategy. And the absence of a strategy, in a system designed to exploit the absence of strategy, was its own form of vulnerability.
He picked up his phone. He called Huang.
"I want to testify," he said.
"To what?"
"If they hold hearings on the bill. Senator Hartwell's bill. I want to testify."
Silence.
"Marcus, you are currently the defendant in three civil lawsuits. Anything you say in testimony could be used in discovery. The Ohio case. The Delaware case. The New York case. Your testimony would be transcribed, entered into the congressional record, and immediately subpoenaed by opposing counsel in all three jurisdictions."
"I know."
"Your testimony could increase your exposure. Significantly."
"I know."
"Then why?"
Marcus thought about this. He did not think about it the way Elena would, analytically, mapping variables and outcomes. He did not think about it the way Kessler would, systematically, calculating costs and benefits. He thought about it the way he thought about most things: as a feeling that preceded any justification for the feeling.
"Because the footnote has a name," he said. "And somebody should say it."
Huang was quiet for three seconds.
"I'll contact the committee staff," she said. "Be ready for the cost."
"I'm already paying the cost."
"You'll pay more."
"I know."
He hung up. He looked at the wall. He still did not hang anything on it. But he stood up, and that was something.