Hackers and scammers have always exploited the anonymity in blockchain technology for crimes like fraud and money laundering.
According to a new study, a new AI model developed by Elliptic can now analyze blockchain transactions and identify money laundering patterns.
Elliptic's model examines massive amounts of transactions and then differentiates between valid and suspicious activity.
The AI model focuses on connections between transactions (subgraphs) rather than individual transactions.
Elliptic’s tests show that the model has a high success rate, indicating the possible launch of a product soon.
In the past few years, the fight against crypto-related crime has become ever more brutal.
Hackers and scammers have devised new tactics to steal and launder billions of dollars on a daily basis.
However, just when it seems hardest to track and go after crypto cybercriminals, AI has entered the fray, with a new way to monitor transactions on the blockchain and identify money laundering attempts.
Let’s go over the recent report from Elliptic and the MIT-IBM Watson AI Lab on how AI can be a game-changer in cracking down on crypto fraud.
One of the main advantages of blockchain technology is anonymity. This makes it easy for users to conduct transactions and send money around, without fear of being monitored.
However, this feature of decentralized public ledgers is also a major problem.
While the anonymity of blockchain makes it easy for harmless transactions to go through, a barrage of criminal activity can also pass, unchecked.
Everything from drug deals to money laundering and fraud is not only possible but also much, much easier.
However, in recent developments, AI is about to emerge as a powerful tool for law enforcement agencies.
Elliptic, AI and blockchains
Elliptic notes that by employing deep learning techniques, AI can be used to analyze massive amounts of transactions on the blockchain, and check for patterns that signify money laundering.
Elliptic’s report shows that researchers used a deep learning AI model to filter thousands of Bitcoin transactions in real time, attempting to differentiate between real and suspicious ones.
Bitcoin subgraphs
The report shows that the AI models used were trained to identify what they call ‘subgraphs’, or “chains of transactions that represent bitcoin being laundered”.
In essence, it checks for relationships between transactions, rather than the individual transactions themselves.
This merger between Elliptic and MIT-IBM has been in the works since 2019 and has yielded "promising results" so far.
The report also mentions a test run of the AI model with an anonymous crypto exchange.
During the test run, the model was able to successfully identify an impressive 52 subgraphs that were likely to be related to money laundering activity.
As confirmation, the exchange in question confirmed that 14 of these subgraphs were indeed linked to users who had been flagged for money laundering-related crimes.
This indicates a pretty high success rate, showing that a real-world product could be launched soon.
This breakthrough might be the first real step towards using AI to mitigate crypto-related crimes, starting from money laundering.
And as the AI landscape continues to change, we might be entering a new era of secure crypto transactions, fewer instances of unnecessary regulation from watchdogs, and a generally safer environment for crypto users.
Disclaimer: Voice of Crypto aims to deliver accurate and up-to-date information, but it will not be responsible for any missing facts or inaccurate information. Cryptocurrencies are highly volatile financial assets, so research and make your own financial decisions.