Local Man Becomes Leading Voice in Anti-AI Movement While Secretly Using AI for Everything

LOCAL — A man who has emerged over the past eighteen months as one of the most recognizable public critics of artificial intelligence — a fixture of panel discussions, opinion pages, and what one organizer described as “the kind of conference that has a lanyard budget” — has reportedly been using artificial intelligence extensively, and without apparent reservation, to accelerate nearly every aspect of his own professional output. The discrepancy, first noted by colleagues and later confirmed by investigators, has been characterized by analysts as the clearest documented instance to date of a strategy now circulating in certain advisory circles under the name Competitive Misinformation Arbitrage.

CLASSIFICATION: COMPETITIVE MISINFORMATION ARBITRAGE ANALYSIS
DISTRIBUTION: Strategy Officers, Thought Leadership Coordinators, Anyone Currently Drafting A Post About How They Drafted It Themselves
PREPARED BY: The Externality Research Division, in consultation with the Institute for the Study of Stated Versus Revealed Preference
DATE: June 2026

The revelation came after a routine review of the man’s working files, undertaken — according to sources familiar with the matter — for unrelated reasons, and abandoned almost immediately once the reviewers encountered the contents. Among the materials reportedly recovered were artifacts consistent with AI-assisted research, AI-assisted writing, AI-assisted planning, and AI-assisted scheduling, as well as a single folder, located at the root of the man’s primary drive, labeled Definitely Not AI Stuff.

Investigators noted that the folder was, by a substantial margin, the largest on the device, and that its contents were “entirely AI stuff.”

The Public Position

In his public-facing capacity, the man has been, by all available accounts, consistent. Across an estimated forty-one recorded appearances reviewed for this report, he has described artificial intelligence, variously, as dangerous, irresponsible, a threat to society, and a force actively engaged in the ruin of entire industries. His delivery has been described by attendees as “urgent,” “sincere,” and, by one repeat audience member, as “one of the strongest anti-AI voices I’ve seen.”

The man’s rhetorical range is, the report notes, narrow but well-maintained. A linguistic analysis of his public statements identified a recurring four-part structure — warning, lament, anecdote, and a closing appeal for collective restraint — that he has deployed, with minor variation, at venues ranging from a regional library to a keynote stage with a fog machine. The fog machine, the report observes, was “not requested in connection with a warning about the dangers of automation, but was present regardless.”

“We are sleepwalking into a future we did not design and cannot control. I want everyone in this room to think very carefully about the tools they are reaching for.”
— The man, addressing a sold-out audience, from notes investigators later determined had been drafted in four seconds

Audience members interviewed afterward described the remarks as moving. Several indicated that they had resolved, on the spot, to reduce their own reliance on automated tools. One attendee reported that he had gone home that evening and deleted an application from his phone. Researchers were unable to confirm whether the application was reinstalled, but noted that this is “the typical trajectory.”

The Private Position

Privately, sources indicate, the man has integrated artificial intelligence into what one associate described as “the entire pipeline, end to end, including the parts that warn people about the pipeline.” The associate, who requested anonymity on the grounds that he too maintains a folder, said that the man’s output had roughly tripled over the period during which his public skepticism had sharpened to its present edge.

Forensic review of the man’s working materials reportedly produced a usage profile that investigators characterized as “total.” Selected findings, reproduced here with the qualification that they are drawn from a single subject and are “not representative of the broader commentariat,” include:

• Share of research conducted with AI assistance: An estimated 94 percent. The remaining 6 percent was, investigators noted, “research into how to phrase his objections to AI more persuasively,” which was also conducted with AI assistance, and was therefore reclassified.
• Speeches warning against AI that were themselves drafted using AI: All of them. The report notes that one such speech contained the line “there are some things a machine simply cannot write,” which had been written by a machine, and which the man delivered without incident.
• Calendar events created, rescheduled, or triaged by an automated assistant: 100 percent. Investigators noted that even the man’s “digital detox weekend” had been scheduled, optimized for travel time, and protected with a recurring hold by the assistant in question.
• Folders labeled to suggest the absence of AI: One. Its contents are described elsewhere in this report.
• Instances in which the man asked a model to “make this sound less like AI wrote it”: 312. The report notes that this figure is “itself a form of AI usage,” a determination the man is said to dispute on grounds he has declined to specify.

When the materials were first presented to the man for comment, he reportedly appeared confused by the question, and asked the investigators to clarify what, specifically, they believed the contradiction to be.

The Explanation

At a hastily organized press conference, convened in a hotel function room that one attendee described as “clearly booked that morning,” the man offered what observers described as a remarkably calm account of his thinking. He did not, witnesses noted, dispute any of the underlying facts. He disputed, instead, that the facts constituted a problem.

“When a competitor is doing something clearly stupid, you’ve got to support them and focus on your bread.”
— The man, at a press conference, in response to a question about hypocrisy

Pressed to elaborate, the man reportedly leaned into the microphone and added a second sentence, which has since been reproduced in at least three business-school case studies and one motivational poster of uncertain authorization:

“Who am I to interfere with a strategic mistake?”
— The man, declining to consider the possibility that the mistake was his own

The man went on to explain that he regarded the public abandonment of AI tools by his peers not as a threat to be argued against, but as an opportunity to be encouraged. Every competitor who took his advice, he indicated, was a competitor who would now be doing the work more slowly, by hand, and at greater expense — a development he characterized as “none of my business, and also the entire point.”

The Strategic Framework

According to documents obtained by reporters — documents the man confirmed he had organized using an automated assistant — the man maintains a private taxonomy that divides all public discourse about technology into two categories. The first is labeled Things People Say. The second is labeled Things People Actually Do. A handwritten note in the margin, the only handwritten artifact recovered during the entire review, reads: do not confuse these.

The man is said to regard the maintenance of a clear separation between the two categories as the central discipline of his professional life. His public statements populate the first category. His working files populate the second. The two are, by design, never permitted to meet, except in the folder labeled Definitely Not AI Stuff, which the report describes as “the seam where the strategy is most visible, and least examined.”

“The mistake people make is thinking the two columns are supposed to match. They are not supposed to match. If they matched, there would be no edge.”
— The man, explaining a framework he insists is “just common sense”

The report notes that the man does not regard this arrangement as deception, on the technical ground that he has never claimed to do the things he recommends. He has only recommended them. The distinction, he is said to maintain, is “the whole job.”

Industry Reaction

Analysts contacted for this report described the man’s approach with a mixture of professional alarm and, in several cases, undisguised admiration. The strategy — publicly encouraging a behavior one privately avoids, in order to disadvantage those who follow the encouragement — has been formally designated Competitive Misinformation Arbitrage in at least two recent strategy notes, one of which was, sources confirm, generated with AI assistance and warns extensively about the dangers of AI-generated content.

One strategist, asked to summarize the man’s position, declined at first to characterize it as adversarial.

“He’s not fighting the anti-AI movement.”
— Strategist, after a pause

A pause followed. The strategist then continued.

“He’s investing in it.”
— The same strategist, who then asked whether the interview was being recorded by anything automated

A competing analyst, asked whether the strategy was sustainable, indicated that its sustainability depended entirely on a single variable: the continued willingness of the man’s peers to take public advice at face value. So long as that willingness held, the analyst said, the man’s position was “not only durable, but compounding.” The analyst then disclosed that she herself had recently announced a return to handwritten correspondence, and declined to say whether the announcement had been drafted by hand.

The Mechanics of the Arbitrage

The report devotes a chapter to the underlying economics, which it describes as “simpler than the terminology suggests.” The arbitrage, the authors explain, exploits the gap between two prices: the social price of saying a thing, and the practical price of doing it. The man, they note, has discovered that the two can be held apart indefinitely, provided one is willing to pay the first price loudly and the second price quietly.

In the man’s case, the social price of opposing AI is paid in front of audiences, and is reimbursed in the form of credibility, speaking fees, and what the report calls “the warm regard of people who will never see his folder.” The practical benefit of using AI is collected privately, in the form of output produced faster and cheaper than that of the competitors who heeded his warnings. The man, the report concludes, “is paid to discourage the very advantage he is busy accumulating.”

Economists consulted for this report were divided on whether the arrangement constitutes a market failure or a market working exactly as designed. One described it as “an information asymmetry the subject is actively manufacturing.” Another described it as “Tuesday.” A third declined to comment, citing a conflict of interest he characterized only as “structural.”

The Question of Credibility

When the press conference concluded, witnesses reported that the man gathered his materials, thanked the room, and was observed — within what one attendee estimated as ninety seconds — opening multiple AI tools on a laptop in the corridor outside. He did not, the attendee noted, appear to be attempting to conceal this. He appeared to be catching up.

A reporter who followed him into the corridor asked whether the open and immediate use of the tools he had spent the preceding hour condemning might undermine his credibility. The man, witnesses said, considered the question seriously before answering.

“Absolutely.”
— The man, without breaking eye contact

Another pause followed. The man then completed the thought.

“That’s what makes it effective.”
— The man, before returning to the tools

The reporter, in a note appended to the transcript, indicated that she had attempted, for several minutes, to locate the flaw in this reasoning, and had been unable to. She further noted that she had filed the present account using an automated transcription service, and that she was “not currently in a position to throw stones.”

The Movement’s Response

Leaders of the broader anti-AI movement, contacted for this report, expressed what one organizer described as “disappointment, but not surprise.” Several reaffirmed their commitment to the cause and indicated that the man’s conduct, far from discrediting the movement, demonstrated precisely the kind of corruption the movement existed to oppose. None, the report notes, asked to have him removed from the speaking circuit, on the apparent ground that he remained “the most effective draw they had.”

One coalition issued a formal statement condemning the use of AI by public figures who oppose AI. The statement, investigators determined, contained stylistic markers consistent with AI generation, including a passage that began “in an era increasingly defined by.” Asked about the markers, a spokesperson for the coalition said the statement had been “reviewed by a human,” and declined to specify the duration of the review.

At least one rival speaker, sensing an opening, announced that he would henceforth conduct all of his work without automated assistance of any kind, and invited the public to hold him to it. Analysts noted that this announcement was, from the man’s perspective, “the best possible outcome,” and that the man had reportedly sent the rival a brief, encouraging message expressing his support. The message, sources confirmed, was drafted by an assistant.

The Closing Statement

The man has vowed to continue speaking against artificial intelligence. He has, sources indicate, already booked his appearances through the end of the year, using a tool he has publicly described as “part of the problem.” His next keynote, provisionally titled The Human Cost of the Machine, is reportedly being outlined this week, and is said to be “coming along quickly.”

When the present report was put to him in full, the man read it — or, witnesses clarified, had it summarized for him by a tool — and offered a single comment. He said that he found the analysis “basically correct,” that he intended to take no corrective action whatsoever, and that he regarded the publication of the report as, on balance, “good for the brand.”

At press time, he remained a respected voice in the anti-AI community. He also remained, by every available measure, an unusually productive member of it.

Update: Following the publication of this report, three additional public critics of artificial intelligence issued statements distancing themselves from the man’s conduct. All three statements, when reviewed, were found to contain the phrase “in an era increasingly defined by.”

Editor’s Note: The Externality maintains no folder labeled Definitely Not AI Stuff. The Externality maintains several, distributed across multiple drives, for reasons of operational resilience.