Helan Singh
February 01, 2024
Procuring poor-quality wine masquerading as premium poses a significant challenge for sommeliers and the wine industry, leading to a lucrative but illegal trade estimated to cost European sales nearly US$3.2 billion in revenue.
Scientists have harnessed Artificial Intelligence (AI) to combat wine fraud by developing an algorithm that detects fraudulent wines by employing chemical analysis to trace the drink's origin.
The algorithm utilises gas chromatography to analyse 80 wines harvested over 12 years from seven Bordeaux estates, creating a unique signature for each premium liquor by drawing on all chemicals detected in the drink.
Rather than focusing on individual compounds, the algorithm groups wines with similar signatures, allowing researchers to differentiate wines based on the subtlest differences in the tiniest chemical compounds.
The AI algorithm successfully revealed chemical clusters corresponding to specific chateaux, effectively mapping Bordeaux based on the chromatography findings and providing insights into the wine's origin.
The algorithm's application demonstrated its ability to scrutinise grapes, soil, and microclimate, providing near-accurate results in determining whether a wine is a vintage, showcasing its potential in preventing wine fraud.
Beyond fraud prevention, the AI program can monitor winemaking quality and technique, ensuring proper blending and craftsmanship, thereby reducing costs associated with producing quality blends.
By confirming whether a wine matches its label, the algorithm becomes a crucial tool in curbing fraudulent networks dealing in fake wines, potentially safeguarding the economic interests of the European wine market.