The development of the Smart ID Engine 2.5 reflects a growing trend in leveraging AI for enhancing security measures in financial services.
The dual architecture approach of using both Da Vinci and Sherlock allows for a comprehensive verification process, addressing both speed and accuracy in document recognition.
The ability to explain AI decisions is crucial in building trust with users, particularly in sensitive areas like financial transactions.
As financial institutions increasingly adopt AI technologies like Smart ID Engine 2.5, we may see a significant reduction in document fraud cases.
The success of this system could lead to wider applications of AI in other sectors requiring secure identity verification, such as healthcare and government services.
Future iterations of the technology may incorporate more advanced machine learning techniques to further improve accuracy and efficiency in document verification.
Russian IT company Smart Engines has developed a neural network system called Smart ID Engine 2.5, designed for remote identification and document verification. This innovative technology aims to combat fraud in financial institutions by accurately recognizing and verifying documents from photographs, scans, and video streams.
The system utilizes two AI architectures, named Da Vinci and Sherlock. Da Vinci is responsible for quickly identifying documents regardless of their orientation, while Sherlock employs a scientific-deductive approach to verify authenticity, capable of detecting both simple and complex forgeries, including those created using deepfake technology.
Sherlock conducts 533 checks divided into 13 categories, analyzing digital authenticity, image integrity, and consistency of details. It can identify fake documents from any country, supporting over 4,573 document templates worldwide. The AI also provides explanations for its conclusions, enhancing transparency in the verification process.
Smart Engines emphasizes compliance with legal standards in AI training and operation, ensuring that real documents are securely stored by customers and not shared with third parties. The company aims to simplify the verification process while strengthening security against fraud. CEO Vladimir Arlazarov highlighted the importance of creating an AI that offers interpretable results to reduce fraud risks without denying service to clients without valid reasons.
