How AI Helps the Enterprise Collaborate on Sensitive Data

More high-quality data is being produced today than at any other period in history. This data, in vast and incomprehensible amounts, is being produced by businesses and individuals. Unlocking it has the potential to positively transform those same businesses and individuals. There is a problem: ensuring privacy, security and intellectual property (IP) safeguards to protect […]

The post How AI Helps the Enterprise Collaborate on Sensitive Data appeared first on PYMNTS.com.

More high-quality data is being produced today than at any other period in history.

This data, in vast and incomprehensible amounts, is being produced by businesses and individuals. Unlocking it has the potential to positively transform those same businesses and individuals.

There is a problem: ensuring privacy, security and intellectual property (IP) safeguards to protect against data leaks or compliance violations.

“Companies and enterprises are increasingly facing a dilemma between how much they want to leverage their data versus how much they want to keep it secure and protected,” Sadegh Riazi, CEO and founder at Pyte, told PYMNTS during a discussion for the “AI Effect” series.

Balancing data security with the need for data sharing is especially pronounced in financial services, where data protection and collaboration are both paramount.

After all, the financial sector handles sensitive information, including personal data, transaction histories and financial records, which if compromised can lead to severe consequences such as identity theft, financial loss and reputational damage.

“You want to make more use of your data to advance the business and do more innovation,” Riazi said. “On the other hand, the more you expose it to, whether it is your internal employees as well as external collaborators, then there’s more risk to the data.” 

But as more companies recognize the value of privacy and security in data collaboration, techniques like secure multiparty computation (SMPC) — a cryptographic way of protecting information that allows businesses to operate directly on encrypted data without the need for decryption — are emerging to allow multiple parties to work together on their data without revealing any sensitive information.

Data Security and Data Sharing

SMPC technology ensures that data remains encrypted at all times, even during analysis. The implications of such security measures are profound. By maintaining encryption throughout the data lifecycle, financial institutions can protect against unauthorized access and breaches. 

“The state of data exchange today is very limited in scope” due to the threats, Riazi said, explaining that Pyte, short for Private Byte, aims to make data collaboration as secure and private as possible using SMPC.

And while data security is essential, data sharing is equally critical for driving innovation within the financial sector. The ability to access and analyze datasets enables financial institutions to enhance services, improve risk assessment, and detect fraud more effectively.

Riazi underscored that limited data exchange hampers the ability of financial services players to detect fraud and accurately assess risks. Financial institutions often operate in silos, with restricted views on customer activities, leading to suboptimal fraud detection and risk management. By enabling secure data sharing, institutions can collaborate without compromising data privacy.

For example, banks could share transaction data to flag suspicious activities without disclosing sensitive customer information. This collaborative approach, Riazi said, can drastically improve the accuracy and efficiency of fraud detection. Similarly, insurance companies can leverage shared data to better estimate risks, leading to more accurate pricing and reduced costs for customers.

Building a Secure Data Ecosystem

Despite the clear benefits, data collaboration within and between organizations presents several challenges. These include regulatory compliance, data sovereignty issues, and internal data silos within multinational corporations. The challenge is even more pronounced when dealing with cross-jurisdictional data flows, where moving data across borders can be legally and logistically complex.

“It’s not ideal to move U.S. data to EU or vice versa. So, you are usually stuck running high-level analysis in each jurisdiction and only being able to combine the final results,” Riazi said.

But by leveraging SMPC solutions, organizations can conduct holistic analyses on encrypted data, ensuring compliance with local regulations while maintaining data privacy. He explained that this approach allows multinational enterprises to gain comprehensive insights without the need to transfer data across jurisdictions.

The rise of artificial intelligence (AI) has further accentuated the importance of data sharing. AI models thrive on large, high-quality datasets to deliver accurate and reliable results. However, the need for data often clashes with privacy concerns, creating a dilemma for financial institutions.

As Riazi highlighted, by using technologies like SMPC, institutions can “rent out” their data for model training without relinquishing control or exposing sensitive information. This method provides a way to improve AI models while maintaining stringent data protection standards.

To realize the full potential of secure data sharing, financial institutions must invest in creating a robust data ecosystem. This involves not only adopting technologies like SMPC but also fostering a culture of security and collaboration within the organization.

Education and awareness are key components of this process. Many in the financial sector are still unfamiliar with secure computation technologies and their benefits. As Riazi noted, part of Pyte’s mission is to educate the market and build confidence in these technologies, demonstrating their practical applications and security advantages.

The post How AI Helps the Enterprise Collaborate on Sensitive Data appeared first on PYMNTS.com.