As enterprises race to feed proprietary data into AI systems, enterprise data security has become the defining bottleneck — not compute, not models, but the sensitive information companies are too cautious to touch.

The challenge is especially urgent for regulated industries, where decades of accumulated data sits locked away in mainframes and unstructured files — emails, PDFs, transcripts — which traditional security tooling cannot safely expose for analytics or machine learning. Turning that “dark data” into a usable asset without compromising compliance is now the central problem of the AI era, according to Prashant Prahlad (pictured), senior vice president of product at Capital One Software, a division of Capital One Financial Corp.

“There’s a false choice presented when it comes to proprietary information or sensitive data,” Prahlad said. “Either you protect it and create a data silo and make sure that only certain people have access to it, or you just allow AI to run loose with it … What we’ve been doing at Capital One for a while is replacing any kind of sensitive information … with format-preserving tokens. That allows your downstream applications to do analytics … without exposing your private information to those applications.”

Prahlad spoke with theCUBE’s Dave Vellante at the RSAC 2026 Conference, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed enterprise data security, tokenization as a foundation for AI readiness and Capital One Software’s expanding product portfolio. (* Disclosure below.)

Enterprise data security through tokenization and dark data unlock

The appeal of tokenization extends well beyond compliance offloading. By replacing sensitive values with mathematically irreversible, format-preserving tokens, organizations can collapse the compliance chain that follows sensitive data through every downstream system, Prahlad explained. An application that never sees a raw credit card number is no longer subject to the full weight of payment card industry compliance, for example, freeing development teams to build on top of that data without regulatory drag, he added.

“If I remove the personal information from that data, the application downstream — as long as it’s not mathematically reversible — doesn’t have to be subject to the same compliance regimes that it did before,” he said. “That unlock allows customers to actually just go use the sensitive data, which is now tokenized, to do really wonderful things. It unlocks innovation.”

At RSAC 2026, Capital One Software announced expanded Databolt capabilities targeting unstructured data, the category that accounts for 70% to 90% of organizational data, and represents the richest source of proprietary signal for enterprise AI. Through automated workflows, Databolt will scan, classify and tokenize sensitive elements inside PDFs, emails and transcripts, making previously dark data safe for retrieval-augmented generation and other AI use cases, Prahlad noted. The solution is also post-quantum safe and carries patents on its core algorithms, he added.

“What differentiates a business long term is the things that are proprietary, that make your models and your AI much smarter than your competitors,” Prahlad explained. “Common knowledge is going to be everywhere. That proprietary data actually exists and … companies have built that over several decades [and now it’s] sitting somewhere because somebody is afraid that this data has sensitive information or the data is not being utilized correctly, or the fear of AI using this data to do something malicious. For us, that’s the opportunity.”

More broadly, Capital One Software is positioning itself as an enabler for safer, AI-enabled enterprise data security. Its core bet is that AI gets better when companies have tools to safely put sensitive data to work, according to Prahlad.

“You want to be a tool builder, right? At the end of the day human beings have become more productive because we have tools that make us more productive. They give you 10x productivity,” he said. “We are in the business of building tools that make the AI better.”

Stay tuned for the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of the RSAC 2026 Conference.

(* Disclosure: Capital One Software sponsored this segment of theCUBE. Neither Capital One Software nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE