AI has learned to detect money laundering through bitcoin

🙀 Elliptic company has developed an artificial intelligence-based system capable of efficiently detecting 🧺 money laundering through bitcoin transactions.

📉 The team used a dataset of nearly 200 million transactions on which the model was trained to identify not just individual illegal operations, but entire "subgraphs".

⌨️ System testing on an unnamed cryptocurrency exchange showed high accuracy: out of 52 suspicious transaction chains, 14 were confirmed as related to money laundering. 🧐 On average, only one in 10,000 accounts is marked as such, indicating the model's effectiveness.

🤯 Using external data, the system can detect money laundering schemes that are inaccessible to traditional analysis methods.

The model enabled the identification of a number of previously unknown wallets used by entities associated with Ponzi schemes, darknet markets, and other illegal activities.