The art authentication crisis is no longer a specialist worry; in 2025 and 2026 it is the central problem a contemporary collector confronts before bidding. Over fifteen years the artist boards that once served as the market's court of last resort have closed, one after another, and no institution has taken their place. Now artificial intelligence generates convincing fakes and, more quietly, fabricates the provenance paperwork that used to reassure a buyer. The result is a market where no committee will vouch for a work and the documents around it can no longer be trusted at face value. This article sets out why the boards went dark, what replaced their verdicts, and the practical way a serious buyer verifies a contemporary work today.
Key takeaways:
- The Andy Warhol Art Authentication Board wound down in 2011 and dissolved in 2012; the Basquiat Authentication Committee and the Keith Haring Foundation's committee both disbanded in September 2012, according to the Center for Art Law.
- The Warhol board faced roughly ten lawsuits over fifteen years, won every one, and still spent more than $10 million on legal defence, according to the Andy Warhol Foundation.
- Authenticity now rests on a due diligence stack: provenance, catalogue raisonne inclusion, documentary evidence, scientific analysis and condition.
- AI now drafts fake provenance letters, invents exhibition histories and seeds online coverage, as ArtsWave reported on 2 March 2026, while some specialists estimate roughly 20% of museum holdings may be forgeries or misattributed.
- Salvator Mundi sold for $450.3 million at Christie's on 15 November 2017 and remains contested, the clearest emblem of how far expert opinion can diverge on a single object.
- The global art market turned over an estimated $59.6 billion in 2025, up 4%, with public auction sales rising 9% to $20.7 billion, according to the Art Basel and UBS Global Art Market Report 2026.
Why the art authentication crisis reached a head
An authentication board is a committee, usually run by an artist's foundation or estate, that issued binding opinions on whether a given work was genuine. For decades these boards were the nearest thing the market had to a final word. Between 2011 and 2012 the most important of them shut down. The Andy Warhol Art Authentication Board, which had operated since 1995, announced in 2011 that it would close and dissolved in 2012. The Basquiat Authentication Committee, formed in 1993, ceased accepting submissions in September 2012. The Keith Haring Foundation disbanded its committee in the same month, as the Center for Art Law recorded at the time. The Roy Lichtenstein Foundation and the Noguchi Museum stepped back from the same work.
For the collector, the closure changed the burden of proof overnight. Under the old system a buyer could outsource the hardest question to a committee and treat its answer as final. The art authentication crisis handed that question back, unanswered, to everyone who buys. A generation of collectors who came of age expecting a definitive yes or no now has to build and weigh the case themselves, or rely on a house that does it for them. That is the quiet structural change beneath the headlines about fakes, and it is why the closures matter far more than any single disputed picture.
The cause was liability, not indifference. A board that declared a work fake could expect a lawsuit from the disappointed owner, whose asset had just lost most of its value. The Andy Warhol Foundation has stated that its board faced roughly ten legal actions over fifteen years, won every one, and still spent more than $10 million defending them. When the cost of being right approaches the cost of being wrong, the rational institution stops issuing opinions altogether. That is precisely what happened, and it is why the authentication boards disbanded across a narrow window rather than one isolated case.
Their closure removed something specific from the market: a single address a buyer could write to for a yes or a no. The International Foundation for Art Research (IFAR), a New York non-profit that runs an authentication research service and publishes provenance guidance, continues to offer scholarly opinions on some artists, but it does not replace the artist-specific boards, and it is careful to frame its work as research rather than a warranty. The art authentication crisis is, at bottom, the story of that missing address and the improvised system that has grown up around its absence.
What a committee's verdict once certified
A board's opinion was never a laboratory proof; it was an expert consensus, delivered by people who had spent careers with the artist's work. What it certified was connoisseurship, the trained judgement that lets a specialist recognise an artist's hand through close visual analysis of brushwork, line, and construction. At its best that judgement was formidable. Its weakness was that it was subjective, hard to audit, and impossible to appeal once the board dissolved.
When the boards vanished, the burden shifted from a single verdict to a body of evidence. A contemporary buyer no longer asks whether a committee blessed the work. The buyer asks whether the evidence, taken together, supports the attribution and survives scrutiny. That is a harder question, and a more honest one, because it forces the weaknesses of each piece of proof into the open rather than hiding them behind an institutional signature.
The art authentication crisis is often described as a loss, and in one narrow sense it is: the market gave up a convenient shortcut. Seen more precisely, it exchanged a fragile certainty for a durable process. A committee could be wrong once and take a reputation or a fortune down with it, with no route of appeal once it dissolved. A well-kept evidence file can be revisited, corrected and strengthened as new information appears, which is the more honest way to treat a question that was never fully closed to begin with.
The practical consequence is that no serious house today rests an attribution on a single document. LLB Auction treats each lot as a case to be built from independent strands of evidence, and treats the condition report as law rather than marketing. Understanding the difference between a certificate of authenticity and a catalogue raisonne entry is the first step in reading that evidence correctly, because the two carry very different weight.
The art authentication crisis rewards this shift from verdict to evidence, because a case built from independent strands can be tested, challenged and priced, where a single opinion could only be accepted or refused. A buyer who internalises the change stops hunting for the one document that settles everything and starts reading the whole file. That habit of mind, more than any single tool, is what separates a collector who navigates the art authentication crisis from one who is quietly exposed by it.
The due diligence that replaced the board
The art authentication crisis did not leave a vacuum; it left a method. What replaced the committee is a stack of five independent checks, none sufficient alone, each capable of exposing a problem the others miss. Provenance, the documented chain of ownership from the artist's studio to the present, is the spine. A catalogue raisonne, the scholarly inventory of an artist's complete accepted output, tells you whether specialists have already recorded the work. Primary documentary evidence such as invoices, exhibition labels and correspondence corroborates the chain. Non-destructive scientific analysis reads the physical object. And the condition report, a professional account of the work's material state, records every repair and alteration that might disguise or reveal its history.
The stack has hard limits, and honesty about them is part of the discipline. Many contemporary artists have no completed catalogue raisonne, so inclusion cannot be checked; Jean-Michel Basquiat is the emblematic case, which is one reason his market has been so exposed to forgery. Provenance often carries gaps, especially for works that changed hands privately. Scientific testing can prove a material is anachronistic, but it rarely proves authorship on its own. The buyer's task is to weigh a case, not to collect a guarantee.
| Evidence type | What it establishes | Where it falls short |
|---|---|---|
| Provenance chain | Ownership from studio to present | Private sales leave gaps; documents can be forged |
| Catalogue raisonne | Scholarly acceptance of the work | Many living or recent artists have none |
| Documentary evidence | Invoices, exhibition and loan records | Paperwork is now the easiest thing for AI to fake |
| Scientific analysis | Materials consistent with the date claimed | Proves anachronism, rarely proves authorship |
| Condition report | Repairs, losses and alterations | Describes state, not authenticity, by itself |
Reading these strands against one another is the working method behind every credible attribution. The attribution ladder that separates an autograph work from a studio piece or a follower's copy is the vocabulary specialists use to grade exactly how strong that combined case is.
None of this makes verification easy, and the art authentication crisis is partly a crisis of effort: the evidence-based method demands hours of cross-checking that a committee letter once spared the buyer. A short provenance from a named gallery, corroborated by a dated invoice and an exhibition label the institution confirms, is worth more than a thick file of unverifiable claims. Depth of evidence, not volume of paper, is the measure that matters, and it is the measure a serious house applies before a work is ever catalogued.
When experts disagree: Salvator Mundi and the limits of the eye
The deepest problem the crisis exposed is that connoisseurship, the very thing the boards certified, is not stable. Experts of equal standing look at the same object and reach opposite conclusions. No work illustrates this better than Salvator Mundi, sold as a Leonardo da Vinci at Christie's in New York on 15 November 2017 for $450.3 million, the highest price ever paid for a painting. Christie's cited a broad scholarly consensus for the attribution. Yet Matthew Landrus of Wolfson College, Oxford argued the panel was largely the work of Leonardo's assistant Bernardino Luini, the Prado in Madrid later downgraded it, and, as The Art Newspaper recounted in its first-hand account of the sale, more than 60% of the top paint layer had been lost to damage and restoration. Half a billion dollars did not buy certainty.
Salvator Mundi is old-master territory, but the lesson runs straight into the contemporary market, where the same subjectivity operates with far thinner scholarship to steady it. History supplies the warning. Wolfgang Beltracchi sold hundreds of paintings falsely attributed to Max Ernst, Heinrich Campendonk and Andre Derain, using period canvases and pigments to deceive museums and specialists, before his arrest in 2010. Elmyr de Hory produced convincing works in the manner of Pablo Picasso, Henri Matisse and Amedeo Modigliani for decades. Both fooled trained eyes for years, which is the point: an unaided expert opinion, however distinguished, is a probability, not a proof.
The contemporary market inherits this fragility with far less scholarship to steady it. A Renaissance panel has five centuries of study behind it; a work made in the last decade may have almost none, no catalogue raisonne, no settled literature, sometimes not even a living consensus about the artist's hand. That thinness is where the art authentication crisis bites hardest, because there is less accumulated knowledge to catch a convincing fake before it reaches a saleroom, and because the sums involved have grown large enough to reward the effort of making one.
This is why the vanished boards cannot simply be mourned as a lost gold standard. A single committee's verdict concentrated all of that subjectivity into one unappealable judgement. The evidence-based method that replaced it is slower and less satisfying, but it distributes the risk across independent checks, and it makes disagreement visible instead of burying it under an institutional stamp. The art authentication crisis has forced the market to confront a truth the boards had let it avoid: attribution is an argument, and arguments can be wrong.
How AI forges the paper trail
The newest turn in the crisis is that forgery no longer needs a forged painting. Artificial intelligence can generate an image in an artist's style in seconds, but the more corrosive use is documentary. As ArtsWave detailed on 2 March 2026, AI is now used to draft polished provenance letters, build gallery websites with fabricated exhibition histories, invent collector timelines and seed online coverage that manufactures a false sense of legitimacy, and even to support insurance fraud on works that never existed. AI art forgery, in other words, targets the evidence stack itself, and it targets the strand that used to be cheapest to trust: paper.
The damage is asymmetric. A convincing fake provenance letter costs almost nothing to produce and can take an expert days to disprove. A whole apparatus of supporting web pages, once impossibly laborious, can now be spun up around a single fake in an afternoon. When the paperwork is suspect, the physical evidence becomes decisive by default, because pigment, canvas and the visible history of the object are far harder to fabricate than a PDF. The Orlando Museum of Art case shows how the physical object still betrays a fake even when the story around it is polished. In 2022 the FBI seized 25 works exhibited as newly discovered Jean-Michel Basquiat paintings after doubts hardened around them; one was made on a piece of cardboard carrying a shipping typeface the courier confirmed had not been designed until 1994, six years after the artist's death in 1988. No provenance narrative survives an anachronism that plain. In the art authentication crisis, that is the reassurance a buyer can still hold onto: fabricated paperwork can be overwhelmed by a single verifiable physical fact, which is exactly why material evidence has gained weight as documents have lost it.
The specialists estimate that roughly 20% of museum holdings may already be forgeries or misattributed works, a figure that predates generative AI and that the new tools can only push upward.
For a buyer, the practical shift is clear. A provenance document is now a claim to be verified against independent records, not a reassurance to be accepted. Exhibition histories should be checked against the institutions named. Gallery websites should be corroborated through sources that the seller does not control. The art authentication crisis has turned the paper trail from the easiest form of proof into one of the most suspect, and a disciplined collector adjusts the weight given to each strand of evidence accordingly.
Can AI catch what AI makes?
If AI deepened the art authentication crisis, it also offers to help contain it. The same technology sits on the other side of the table. Firms such as Art Recognition, co-founded by Dr Carina Popovici, and Hephaestus Analytical apply machine learning to the authentication problem. Their systems use convolutional neural networks and vision transformers, models that break an image into a grid of patches and compare each against an artist's known patterns, then produce heat maps highlighting areas that match or deviate. Applied well, the tools do three useful things: stylistic analysis of brushwork and composition, material analysis of pigments and canvas, and provenance tracking through large database searches, as Art & Object reported in November 2025.
The limits are as important as the capability. These models need training data, and they struggle with the artists who most need help: those who changed style across a career, and those without a catalogue raisonne to supply reliable examples. Worse, competing models disagree. On the de Brecy Tondo, one system read the face as a 95% match to Raphael while another put it at 85%, a spread wide enough to change a conclusion. Denis Moiseev, chief executive of Hephaestus Analytical, put the position plainly: "AI is not a silver bullet."
The honest reading is that AI authentication is a probabilistic aid inside an arms race, not a replacement for the vanished boards. A model can flag a work for closer study or strengthen a case already built on provenance and materials. It cannot, by itself, confer authenticity, and treating a confidence score as a verdict simply relocates the connoisseurship problem into a black box. The tools are worth having; they are not worth mistaking for proof.
A further weakness is reproducibility. A verdict a buyer can rely on must be repeatable and open to challenge, yet proprietary models rarely disclose how they reach a score, and two systems trained on different examples can split on the same canvas. Until that opacity is resolved, an algorithmic confidence figure belongs in the evidence file as one more strand, weighed against provenance and materials, not placed at the top of the pile. The art authentication crisis will not be solved by the same technology that is helping to deepen it, and a buyer who expects a machine to hand back the certainty the boards once offered has misread the problem.
How to verify a work before you bid
The absence of a board does not leave a buyer helpless. The art authentication crisis has a practical answer, and the answer is a protocol rather than a person. It replaces one verdict with a sequence of checks, and the sequence is teachable. The following sequence is how a disciplined collector, and a house acting as the buyer's advocate, works through a contemporary attribution before any paddle is raised.
- Read the provenance chain for depth and gaps. Trace ownership from the studio forward and mark every break. A short, unbroken chain from a named first owner beats a long one full of anonymous private sales.
- Check the catalogue raisonne. Establish whether one exists for the artist and whether the work is in it. Absence is not proof of a fake, but presence is a strong positive signal.
- Verify the documents against independent records. Confirm invoices, exhibition labels and loan histories with the institutions and galleries named, using sources the seller does not control. In the age of AI art forgery this is the step that most often exposes a fabricated file, because an invented exhibition cannot be confirmed by the museum that never held it.
- Commission non-destructive scientific analysis when the value or the doubt warrants it. Pigment, medium and support can be tested against the date claimed. A material that did not exist when the work was supposedly made ends the discussion.
- Read the condition report in full. Repairs, overpaint and relining change both value and the story a surface tells. Learning how to read a professional condition report is inseparable from verifying the work, and knowing when scientific testing belongs in the room tells you which claims a lab can actually settle.
- Check the stolen-art and restitution registers. Confirming clean title through the checks that protect a buyer against a contested claim is part of authenticity in the full sense, because a genuine work with a defective title is still a problem.
- Consult the foundation or estate where it still offers guidance. Some estates answer research enquiries even after closing formal committees. An informal steer is not a verdict, but it is a data point.
- Weigh the whole case, and price the residual doubt. No single check is decisive; the judgement is whether the strands cohere. Where doubt remains, it belongs in the price, not in a hopeful assumption.
Run in full, this protocol is the working answer to how to verify art authenticity in a market with no board to appeal to. It is more demanding than writing a letter to a committee. It is also more robust, because it never rests the decision on one fragile point of proof. The art authentication crisis raises the cost of that diligence, but it rewards it too: the collector who does the work holds a defensible position rather than a hopeful one.
FAQ: the art authentication crisis
What is the art authentication crisis?
The art authentication crisis is the gap left when the artist boards closed and nothing replaced their verdicts. The Warhol board wound down in 2011 and 2012, and the Basquiat and Keith Haring committees disbanded in September 2012. Authenticity now depends on assembling provenance, catalogue raisonne inclusion, documents, science and condition into a case, at the same moment AI has begun to fabricate the documentary side of that evidence.
Which authentication boards have closed, and when?
The Andy Warhol Art Authentication Board announced its closure in 2011 and dissolved in 2012. The Basquiat Authentication Committee stopped accepting submissions in September 2012, and the Keith Haring Foundation disbanded its committee the same month. The Roy Lichtenstein Foundation and the Noguchi Museum also stepped back from issuing opinions, according to the Center for Art Law.
Is a work without a catalogue raisonne entry a fake?
No. Many contemporary and recent artists have no completed catalogue raisonne, so the absence of an entry proves nothing by itself. It simply means one strong strand of evidence is unavailable and the remaining checks, provenance, documents, science and condition, must carry more of the weight. Basquiat is the clearest example of a major artist without a definitive catalogue.
How worried should I be about AI-generated fakes?
The realistic concern is less the fake image than the fake paperwork. AI can now mass-produce provenance letters, exhibition histories and supporting web pages, which erodes trust in the cheapest form of proof. Treat any document as a claim to be checked against independent records, and give more weight to physical and scientific evidence, which remains far harder to fabricate.
Who authenticates contemporary art now?
No single authority does. Attribution is now built from independent evidence and read by specialists, conservators, scientists and reputable auction houses acting on a buyer's behalf. Bodies such as IFAR offer research opinions on some artists, and estates sometimes answer enquiries, but the responsibility for weighing the case, and the risk, sits with the buyer and the professionals they trust.
Does an auction house guarantee that a work is authentic?
A responsible house does not guarantee authenticity; it documents the evidence and states what that evidence supports. At LLB Auction the authenticity risk sits with the buyer, which is why the house publishes its provenance findings and a full condition report rather than offering assurances. The discipline is to let the documented evidence and the condition report speak, and never to overstate what can be proved.
How LLB Auction verifies a lot before it reaches a sale
LLB Auction is built for exactly the market the art authentication crisis has created: one where a buyer needs the evidence laid out and weighed, not a slogan about quality. The house does not claim to have solved the art authentication crisis, and it is wary of anyone who says they have; it commits instead to doing the verification work in full and showing it. The house works as the buyer's advocate, and its due diligence runs before a lot is ever offered.
Curatorial intake. LLB Auction rejects roughly 40% of submissions at intake, so that a work which reaches a sale has cleared a real evidentiary threshold rather than a marketing filter. The rejection discipline is the buyer's working substitute for the board that no longer exists.
Provenance and authentication due diligence. Every lot receives certificate verification, an ownership-history review and a professional three-page condition report with photography, technical specifications and conservation notes. The house documents the evidence and treats the condition report as law; it never claims certainty and never implies it bears the authenticity risk for the buyer.
Transparent terms. Fees are disclosed before bidding, with a buyer's premium of 20% and no hidden costs, across timed online sales of seven to fourteen days. The Contemporary Art Spring 2026 sale, held on 26 May 2026, offered 25 lots and sold 23, each with its provenance and condition documentation published alongside.
Collectors weighing a contemporary work in a market without a board can consult an LLB Auction specialist on the provenance and condition of a specific lot, and receive the documented evidence to read for themselves before the bidding opens.
Conclusion
The closed boards are not coming back, and no committee will be reconstituted to hand a buyer a verdict. What has replaced them is harder and, on balance, better: a body of evidence, read critically, that makes the strength and the weakness of every attribution visible. The art authentication crisis is genuinely a crisis of trust in documents and in single opinions, sharpened by AI that can fake both, but it is not a crisis without an answer. The answer is method. A collector who reads provenance for its gaps, checks the catalogue raisonne, verifies documents against independent records, tests materials when doubt warrants it, and prices the residual uncertainty is better protected than any board ever made them. In the art authentication crisis, the disciplined buyer, not the vanished committee, is the last line of defence.
A lire egalement :
- Certificate of authenticity or catalogue raisonne: which proof counts
- The attribution ladder: autograph, attributed to, studio of, circle of, follower of
- How to read a professional condition report
- When science enters the room: testing a contemporary work
- Restitution and title: the checks that protect a buyer
Sources :
- Authentication Committees Disband: Warhol 2011, Basquiat 2012, Who's Next? : Center for Art Law, 2012
- Basquiat Committee to Cease Authenticating Works : ARTnews, 2012
- Haring and Basquiat foundations will no longer authenticate works : Antiques Trade Gazette, 2012
- Frequently asked questions : The Andy Warhol Foundation for the Visual Arts
- Beyond the Basics: AI and Art Authentication : Art & Object, 2025
- Artificial Authenticity: Art Forgery and AI : ArtsWave, 2026
- Five years since the 450m dollar Salvator Mundi sale : The Art Newspaper, 2022
- Authentication Research : International Foundation for Art Research (IFAR)
- Art market data and price database : Artnet News