The Future of Big Data Security in Financial Services

  • Riddhiman Das, Co-Founder and CEO at TripleBlind

  • 02.08.2021 11:15 am
  • #data #AI #security

Financial services companies are rich in data.  But the vast majority of the data they store cannot be accessed or commercialised due to privacy concerns, operational complexity and regulations.  Data sharing enables institutions to work with larger and more diverse data sets, which helps ensure higher quality results.

Most solutions that enable and facilitate data sharing are insufficient.  Legal agreements to safeguard data sharing are complex, require lengthy negotiation and rely on a high degree of trust between the parties. Technology solutions can be demanding on compute resources, such as homomorphic encryption, or provide an incomplete solution for enforcing regulatory standards such as GDPR, in the case of secure enclaves.

TripleBlind has built a new cryptographic technique which uses AI and machine-learning to enforce privacy regulations while enabling commercialisation of the data.  Today, to share or commercialise data, financial institutions typically must allow their data to be decrypted which can lead to privacy violations, regulatory compliance issues or intended and unintended data abuses.  Based on an approach called ‘blind learning’, the technology allows only approved operations to be executed on data which remains encrypted and is never decrypted.  Using an API-driven virtual exchange, encrypted data can be shared and used safely by financial institutions without ever leaving the data owner’s firewall.  This is especially important for organisations in Europe eager to share their data with non-European organisations while continuing to enforce GDPR.  The most sensitive information within each data set remains private and enables data sharing without reliance on trusting the recipient.  

For example, banks and fraud detection companies can use the technology to collaborate on encrypted data to aid credit card fraud and other types of financial fraud detection. The technology allows issuers, banks and merchants, even competitive institutions, to collaborate without revealing proprietary information.  The fraud detection company can access spending habits of customers, across different banks and accounts, while the banks only receive the essential information necessary to determine if customer accounts have been compromised.  

The technology can also be used to enable bank vendor data monetisation.  A bank’s vendors only ever interact with encrypted data so an information breach at the vendor does not compromise the bank’s data and all information sharing is for an agreed-upon purpose.  This paves the way for secure collaboration with vendors who are able to license novel algorithms to the bank while the bank controls fine grained permissions on data.

Privacy-protecting technology using AI and machine-learning is crucial for the future of the data economy.  It enables financial institutions to safely share sensitive data, collaborate with competitors, and unlock opportunities to monetise their data, without giving up proprietary data – a win-win for all. 

 

Other Blogs