AI tool claims 97% efficacy in preventing ‘address poisoning’ attacks
Trugard and Webacy have developed an AI-based system to detect crypto wallet address poisoning, achieving a 97% success rate in tests. Address poisoning scams involve attackers sending small amounts of cryptocurrency from addresses resembling a target's, tricking users into copying the wrong address. A study reported over 270 million poisoning attempts on BNB Chain and Ethereum from July 2022 to June 2024, with 6,000 successful attempts resulting in over $83 million in losses. The new system utilizes supervised machine learning, adapting to evolving attack patterns and emphasizing context and pattern recognition. Trugard's approach includes generating synthetic training data to simulate attack patterns and continuously updating the model with new data.
