Intel, FiVerity, and Fortanix team up to launch an AI-driven fraud detection platform into a confidential computing environment.
Dark Reading Staff
April 20, 2022
A new partnership will bring confidential computing to financial services companies in their fight against digital identity fraud.
Confidential computing tackles the challenge of securing data during processing or runtime by using hardware-based trusted execution environments (TEE), an environment that focuses on data integrity, data confidentiality, and code integrity. The data, as well as the application processing that data, are isolated within secure enclaves so that nothing else can access that data, even if the infrastructure is compromised. This way, enterprises can run sensitive applications on public clouds or other hosted environments without worrying about the possibility of data exposure.
The partnership is between Intel, digital fraud prevention platform FiVerity, and cloud security company Fortanix. FiVerity's anti-fraud platform relies on AI to detect fraud and incorporates the latest data privacy technologies. FiVerity will rely on Intel Software Guard Extensions to protect the data and control access, while Fortanix will deliver the confidential computing environment for FiVerity's fraud-detection models and transaction data from financial institutions.
Financial institutions would be able to analyze intelligence on fraud activity without exposing personally identifiable information, keeping banks in compliance with security and privacy requirements and reducing their exposure to digital fraud, the three companies said in a joint statement.
The confidential computing market is estimated to grow to $54 billion by 2026, according to a recent report from the Linux Foundation and Confidential Computing Consortium and conducted by the Everest Group.
"We believe that digital fraud detection is a particularly important and timely challenge for the industry which can be addressed with this technology," Fortanix CEO Ambuj Kumar said in a statement.