Author’s note:
“While the foundational concepts and arguments were conceived independently, artificial intelligence was deliberately and experimentally integrated into the drafting and development process through an iterative human-led dialogue. Throughout this collaboration, the author has maintained active editorial and conceptual control. This approach serves a dual purpose: to evaluate the practical utility of AI in academic writing as an extension of original human thought, and to invite critical reflection from the reader regarding its efficacy. As an open experiment in scholarly methodology, the resulting text remains fully subject to academic scrutiny concerning how effectively AI can enhance not only contextual richness, but the intrinsic quality of essayistic work.”
Cowboyrecherche
vs.
De Slachter van Mons
Independent Cold-Case Investigation Public Sources and Artificial Intelligence
Between 1996 and 1997, the dismembered remains of five women were discovered in and around Mons, Belgium. Despite an extensive criminal investigation involving thousands of leads, witnesses, addresses, vehicles and potential suspects, no perpetrator was conclusively identified. Nearly three decades later, the case of De Slachter van Mons — the Butcher of Mons — remains unsolved.
Cowboyrecherche vs. De Slachter van Mons is a two-volume Dutch-language investigative project by Filip Firlefijn. It revisits the publicly accessible record through digital cartography, open-source intelligence, historical research, geographical comparison and AI-assisted analysis. The project does not claim to replace professional criminal investigation or to solve a murder case through artificial intelligence. Its purpose is to examine whether publicly available fragments can generate more precise, historically testable questions — and to establish where the limits of such an approach lie.
The first volume, Een kaart zonder dader (A Map Without a Perpetrator), explores the wider landscape of the case: industrial and cross-border corridors, waterways, railway infrastructure, the station district, social environments, institutions, possible criminal networks and competing explanations for the geographical distribution of the remains. The investigation gradually moves from broad possibilities toward increasingly specific spatial hypotheses.
The second volume, De insiderlasso, returns to the local field. It examines whether certain recurring locations and road connections around Mons, Cuesmes, Hyon and Havré could belong to an identifiable operational geography. Historical road access, railway barriers, ordinary residential and working environments, vehicle movements and possible storage spaces become central questions. A suggestive pattern is never treated as proof that such a route was actually travelled.
Throughout both books, established facts, secondary reporting, geographical observations, inference, speculation and symbolic resonance are carefully distinguished. AI serves as a research instrument, not an authority: it can help compare material, formulate alternatives and expose inconsistencies, but it can also produce persuasive connections that historical evidence does not support.
The project is therefore more than a reconsideration of one unresolved criminal case. It is also an experiment in public criminology, digital humanities and the formation of knowledge in the age of artificial intelligence.
When do scattered facts become an investigative pattern — and when does an apparently coherent reconstruction remain nothing more than a hypothesis?
Een kaart zonder dader
De Slachter van Mons: Een onderzoek met publieke bronnen en AI
Book 1 of Cowboyrecherche vs. De Slachter van Mons
What can an independent researcher discover about an unsolved murder case when the original police file remains largely inaccessible?
Een kaart zonder dader begins with that limitation. Rather than presenting a new suspect or promising a solution, it reconstructs the publicly known geography and investigative history of the Butcher of Mons case. Newspaper archives, historical maps, open-source information, institutional histories and artificial intelligence are brought into a sustained dialogue.
The investigation moves across different scales. It considers possible French connections, industrial and waterway corridors, earlier disappearances, the social environment of the station district, care institutions, criminal networks and the practical geography of the sites where human remains were discovered. Some hypotheses appear initially compelling but weaken under closer examination. Others reveal overlooked questions about access, timing, local familiarity and the relationship between the victims' social worlds and the places associated with the crimes.
The book also examines the research process itself. Digital maps can make scattered locations appear connected; repeated names and dates can acquire an unexpected force; AI can organise evidence while simultaneously amplifying weak assumptions. The challenge is not simply to identify patterns, but to determine which of them survive historical verification and serious counterargument.
As the inquiry progresses, several geographical lines draw increasing attention toward Mons, Cuesmes and the southern urban corridor. Yet proximity does not establish causality, and a concentration of possible connections does not identify an offender.
The volume ends without claiming a solution. Instead, it leaves behind something more limited, but potentially useful: a structured landscape of competing hypotheses, unresolved contradictions and questions that can be investigated further.
A map has emerged. The perpetrator has not.
De insiderlasso
Van ruimtelijke resonantie naar onderzoekswaarde
Book 2 of Cowboyrecherche vs. De Slachter van Mons
A map has emerged. But what, if anything, can it reveal about the operational world behind an unresolved crime?
De insiderlasso continues where Een kaart zonder dader ends. Having explored a wide range of geographical and historical hypotheses, the investigation returns to a more concentrated local field around Mons. The central challenge is no longer merely to observe connections between locations, but to establish whether those connections could have existed as practical routes, everyday movements and usable spaces during 1996–1997.
Particular attention falls on Cuesmes, Rue Émile Vandervelde, Chemin de l'Inquiétude, Avenue du Général de Gaulle, the N6 and the railway infrastructure separating parts of the southern urban landscape. The inquiry then extends toward Hyon, Saint-Symphorien and Havré, examining whether these areas could have formed a series of connected — or partly overlapping — traffic environments.
The image of the lasso offers a way to explore recurring movements around familiar territory. It is not presented as a mathematical spiral, a proven criminal itinerary or evidence of deliberate geographical design. Indeed, one of the book's crucial questions concerns a possible railway passage near Cuesmes whose existence and vehicle accessibility in 1996–1997 remain unconfirmed. If that connection did not exist, an important version of the local circulation hypothesis must be revised.
The investigation also turns from roads toward possible operational spaces: residential backyards, garages, workshops, storage facilities and the less visible infrastructure of ordinary urban life. The question is not whether a particular building or profession is suspicious, but what historically documented combination of access, mobility and personal circumstances might eventually make a hypothesis worth examining.
Contradictory evidence remains central. The chronology of the discoveries does not support a simple progression from Cuesmes through Hyon to Havré, and geographical resemblance cannot substitute for forensic evidence or an independently established biographical connection.
De insiderlasso is consequently an exercise in methodological restraint as much as geographical investigation. It asks whether an increasingly precise map can produce meaningful investigative tests without turning neighbourhoods, institutions or ordinary users of public space into unsupported suspects.
The objective is not to make the map more convincing. It is to discover which parts of the map can withstand verification.
Overview of Authorities and Institutions Contacted
Correspondence with judicial and law-enforcement authorities in 2026, as the Mons case approaches its anticipated legal deadline in 2027.
Standing Police Monitoring Committee (Committee P) — Brussels, Belgium
Federal Judicial Police (PJF) — Mons, Belgium
Public Prosecutor's Office — Valenciennes, France
Judicial Court (Tribunal judiciaire) — Avesnes-sur-Helpe, France
Federal Police — General Directorate of the Judicial Police (DGJ) — Brussels, Belgium
Public Prosecutor's Office (Parquet de Mons) — Mons, Belgium
Corresponcence with news services.
VRT (Vlaamse Radio- en Televisieomroeporganisatie, or: "Flemish Radio and Television broadcasting organisation", a national public service broadcaster for the Flemish Community of Belgium)
VTM (Vlaamse Televisie Maatschappij, the main commercial television station in Flanders and part of a network of channels owned by DPG Media)