According to the Preventing Financial Crimes Playbook prepared jointly by PYMNTS and NICE Actimize, the use of Artificial Intelligence and Automatic Learning could be applied in financial systems to help prevent the increase of financial cybercrime. Some $217 billion would already have been invested by banks to develop applications of this technology.
As mentioned in the text, with the application of these Bitcoin Profit tools, they can analyze thousands of transactions in order to detect elements that can be related to fraud, such as very large transactions, attempts to log on to devices that are made in regions that are not usual for the user. In fact, as also explained, several banks have already been applying these tools and have announced a decrease in financial crime of up to 50%.
NICE Actimize also mentions that banks could „incorporate all relevant events“ of customers in order to „build common patterns and profiles of their customers“ about their logins and payments. The intention is that, if a transaction is submitted that does not „match“ the profile created, it can be cancelled for being classified as fraudulent.
An Invasion of Privacy?
Certainly, fighting financial crime is an objective that is held within all sectors, however, the approach by the experts of using Artificial Intelligence and Automatic Learning in order to achieve the construction of certain „patterns and profiles“ could lead to the banking sector knowing in depth everything that its users do, which would definitely be an invasion of privacy.
In addition, it is necessary to specify how user information will be acquired. As we reported earlier in Cointelegraph, a Federal court recently ruled that the mass collection of data by the NSA is illegal.
In addition to the privacy violation, it would also imply a limitation on what the user wants to do with his or her money. In the event that the person wants to make a transaction outside of his or her „pattern and profile,“ the bank could block the transaction as a fraud, when in fact it could be a user’s own operation.