Your Fancy Wine Could Well Be Fake. Some Are Hoping to Sniff Out a Solution

We all know you can fake a watch or a handbag, but a wine? It’s a real thing, and much more widespread than you might imagine. Accurate estimates for the scale of the problem are hard to come by, but most experts—as well as bodies such as Interpol and the World Trade Organization—settle on a figure of around 20 percent. That is to say, as much as a fifth of all wine sold around the world could be fake, and some say even that is a conservative number.

The best known frauds are high-stakes cases of audacious—and ultimately overambitious—forgeries, such as those carried out by Rudy Kurniawan, who faked expensive vintages to be sold at auction, only to be rumbled when he forged wines that should never have existed in the first place. But the problem of fake wine spans the entire industry, from Jacob’s Creek to Domaine Romanée-Conti.

Earlier this year, French, Italian, and Swiss police forces teamed up to bust a ring of wine fraudsters who were selling counterfeited Grand Cru wines (the premier designation of quality in French wine) for up to €15,000 a bottle. The gang was estimated to have made more than €2 million from the racket.

Wine fraud can take various forms; culprits sometimes combine lesser wines to resemble bottles of top-tier vintages (as Kurniawan did), or sometimes simply pass off a cheaper wine as a more expensive one by faking the labels, cork and foil. In some instances wines can be diluted or altered with cheaper substances, occasionally even harmful chemicals.

At the top end of the market, a fraudster might only need to create a handful of counterfeit bottles to turn a healthy profit—the best Bordeaux vintages can sell for six figures a bottle—but in other cases the scale can be surprising. In 2022, a Spanish operation was dismantled that authorities said could have passed off as much as 40 million bottles of cheap table wine as finer tipples.

Attempts to combat the issue broadly fall into two camps: prevention and detection. The former focuses on ensuring that the wine you buy is authentic by virtue of tracing its every move from the moment it’s bottled; the latter encompasses a wide range of emerging technologies that try to verify a wine from its physical properties alone.

One company looking to make its mark in the traceability side of things is Crurated, a London-based platform for wine distribution and retail that offers its customers a blockchain-backed system for tracking individual bottles throughout their journey from vineyard to restaurant or customer. For CEO Alfonso de Gaetano, a former Google executive, it arose from his own frustrations as a wine collector.

“I was buying from auction houses and [was] really disappointed about the quality of the bottles that I was opening; I was spending a lot of money and the bottles were not really great. You never know if you have to blame the producer or the auction house; it's difficult because there is no guarantee of provenance.” Gaetano explains that even legitimate wine can suffer. “You don't know how many tours of the world the bottle has done. The more you move the wine, the more the bottle gets ruined.”

Crurated offers winemakers a system wherein they can register bottles as soon as they are produced, associating each one with an NFT and creating a blockchain-backed record of its movements. It operates a warehouse in Burgundy, as well as one in Hong Kong, storing wines both for distribution and on behalf of individual owners, while a new “metaverse”-style interface allows the wine’s owner to see its exact location on a 3D walkthrough of the shelves. Of more practical use is an NFC tag embedded in the foil seal of a bottle which not only enables its journey to be logged, but also registers when the wine has been opened.

“If you try to open the bottle, then the circuit breaks and you know that nobody can take that label and use it somewhere else,” explains Gaetano. Crurated’s challenge, he continues, was building a physical logistics system that integrated seamlessly with blockchain tracking.

“We can do it because we have control over the supply and the demand—we know the producers and the ultimate clients. If you start to add more layers, the problem becomes exponentially complex. We are the only ones able to do this globally because we work directly with the clients and there is no one else in the middle.”

It is a salient point, because the blockchain-backed “passport” for each wine only works if everyone who handles the bottle is signed up to it. For this same reason Gaetano won’t introduce wines into Crurated’s system that aren’t coming directly from the producer, but he says he is considering syndicating out his technology to them.

“One of the best producers in Burgundy was asking us if they could use our labels on all their bottles. Then we can create an additional step so that when you buy the bottle, let's say from the distributor in the UK or in the US, you can add the bottle into your wallet," he says. “But now, the idea is that you take out the bottle only when you want to drink the wine. Otherwise it doesn't make sense because you are removing the bottle from the perfect provenance chain.”

Gaetano admits that, strictly speaking, Crurated’s system doesn’t prevent dedicated fraudsters switching out the contents of a bottle (if they can circumvent the NFC tag in the neck), but says that trustworthy authentication comes from never allowing the wine’s whereabouts to be unaccounted for.

If you do find yourself needing to ascertain whether a wine is the real deal, then you’ll need a different technological solution entirely. Some wineries have employed advanced printing techniques for their labels, embedding holograms and printing with invisible inks, but the real prize is an authentication process for what’s inside the bottle.

The number of different parameters to test for—the age of the wine, its place of origin, its chemical composition—means that the problem has been attacked in different ways. A team from the University of Adelaide was able to demonstrate that absorbance-transmission and excitation-emission matrix (A-TEEM) spectroscopy, essentially a very sophisticated scan of a sample, could reliably ascertain the vintage year of a selection of Shiraz wines, also accurately associating every one with a particular subregion of the Barossa Valley area.

Similarly, different studies have shown that nuclear magnetic resonance (NMR) spectroscopy, which works along similar lines to an MRI scanner, can detect different levels of deuterium, a hydrogen isotope, and different amino acids in wine, enabling scientists to identify different vintages and types.

The terroir of a vineyard can be “fingerprinted” in terms of the rainfall it experiences, with different areas known to have chemically distinct rainwater: A 2007 paper showed that analysis of “stable isotopes” within the water used to make the wine could accurately distinguish between different regions of California and Oregon.

Perhaps surprisingly, even the most renowned experts acknowledge that it can be impossible to detect a fake by smell or taste, no matter how nuanced the palette. But where mankind’s nose is defeated, a machine may still sniff out the truth. A team of academics from multiple institutions published a paper in 2023 that showed that by using a method called gas chromatography to analyze the aroma profiles of 80 Bordeaux wines, they could distinguish between vintages from seven particular estates across the left and right banks of the river.

The olfactory route is one already being taken in other industries, with proven results in detecting inauthentic perfumes and even fake sneakers. Applying it to wine makes obvious sense, says Tristan Rousselle, founder and CEO of French firm Aryballe, a specialist in the field. But it also presents unique challenges.

“The overall composition of wine is mainly water,” says Rousselle. “The ethanol and the volatile compounds that give the aroma will maybe be less than 1 or 2 percent of the composition. Usually in a high-end perfume, the perfume ingredients will make up at least 70 percent of the liquid.”

Aryballe uses yet another phenomenally complicated-sounding process—silicon photonics interferometry—which uses a tiny sample of wine. “We need less than a milliliter,” says Rousselle. “Gas is generated from the liquid in the ratio of 1 to 1,000, expanding as it evaporates. So we could do less if you don't want to damage your product, because corks are slightly porous, you can put a needle inside it without damaging it. We could collect 100 microliters of wine, which would generate 100 ml, which is sufficient to do an analysis.”

The need for a physical sample is an obvious drawback, not just for Aryballe but for nearly everyone attempting to solve the problem. Removing even a tiny volume of wine would destroy the value of a rare vintage at auction, although it might be more acceptable to a distributor batch-testing cases of midlevel wine.

Another drawback is reference data: Crurated is building a robust model for tracking and documenting thousands of bottles, but doesn’t store a chemical signature of the wine. Aryballe, to date, has not commercialized its technology in the wine world, but Rousselle explains that each test would require a reference aroma to compare with the subject bottle. Taking a random wine and attempting to divine its properties “blind” would be “much more difficult,” he says.

Ultimately, a foolproof test for fake wine remains a far-off idea. But the combined weight of scientific endeavor being applied to the problem—plus a healthy dose of old-fashioned common sense—should mean the next Rudy Kurniawan is stopped before they ever really get started.

And in the meantime, it may be a first-world problem of the highest degree, but those at the best tables should be able to know their money isn’t being wasted. “We want to guarantee that when you order a Lafite 96, that's the best possible bottle of Lafite 96 that you can drink,” says Crurated’s Gaetano. Cheers to that.