Across industries, organizations are sitting on vast amounts of information that could improve decision-making, customer understanding, and operational performance. Yet a significant portion of that data remains unused. According to Jackie Peters, CEO and founder of Blind Insight, many executives recognize the potential value in these datasets, but privacy obligations, security concerns, and regulatory requirements frequently limit how that information can be used.
She notes this moment as a strategic inflection point for enterprise leaders. "Companies are caught in a difficult position," she says. "The most valuable data they possess often includes customer behavior, operational patterns, and sensitive records, but those same characteristics make it risky to share or analyze without the right safeguards," Peters notes. Organizations increasingly want to apply artificial intelligence tools to these datasets, yet many remain hesitant because they cannot expose confidential information to external systems or uncontrolled environments.
According to Peters, the tension creates a paradox that many executives now recognize. "If companies do not use the data they collect, they sacrifice insights that could guide innovation and strategy," she says. "If they attempt to use it without sufficient controls, they risk regulatory violations or security incidents." Peters argues that this tension has intensified as AI tools become widely available inside organizations. "Employees see the potential of these technologies and naturally want to experiment with them," she explains. "But without clear frameworks for handling sensitive data, experimentation can introduce new risks."
Research also suggests that the appetite for artificial intelligence is accelerating across large organizations. A recent analysis found that 53% of executives say AI will have transformed business models in their industry by 2030. Yet enthusiasm often advances faster than governance structures. According to Peters, organizations frequently restrict access to sensitive datasets while simultaneously searching for ways to apply AI more broadly across their operations.
Blind Insight was created to address this structural challenge. The company develops privacy-preserving data infrastructure designed to allow organizations to analyze and query sensitive information while keeping it encrypted. From Peters' perspective, the concept emerged from years of working with regulated data in fraud, healthcare and technology environments, where privacy obligations frequently limited collaboration and innovation.
"Organizations have spent years building systems that collect and store enormous volumes of information," she explains. "But the ability to safely mobilize sensitive data has lagged behind. Our goal was to design infrastructure that allows companies to extract value from that information without exposing the underlying records."
Blind Insight's platform focuses on enabling queries, analytics, and machine learning operations directly on encrypted datasets. Instead of decrypting sensitive records before analysis, the system processes queries while the information remains protected. Peters notes that by combining this approach with field level access controls alongside AI powered monitoring and reporting, enterprises can explore insights while maintaining compliance with privacy standards and internal governance requirements.
The company is also developing a capability known as Blind(L)LM, which is designed to enable language models to interact with encrypted enterprise data. Peters explains that the idea is to allow artificial intelligence systems to retrieve aggregate insights without accessing raw confidential records. "The model can generate answers using verified numerical results from encrypted datasets," she says. "That means organizations can benefit from AI-driven analysis without exposing the sensitive data itself."
Building such infrastructure required collaboration with specialists in applied cryptography and large-scale systems engineering. She notes that the founding team includes experienced researchers and security experts. Among them is technical co-founder Nick Sullivan, previously head of research at Cloudflare and a long-time contributor to internet security standards. According to Sullivan, privacy-preserving computing is becoming a critical building block for the next phase of enterprise technology. "Organizations want to work with their most sensitive data but few companies have access to tools that allow them to do so," he explains. "The challenge is creating systems that allow that work to happen without compromising security and making it so simple to use that anyone can use it."
For executives evaluating AI strategies, Peters believes the larger lesson extends beyond any single product or platform. She underpins the importance of privacy by design, where companies view privacy-preserving infrastructure as a foundational component of modern data strategy. "Artificial intelligence is only as powerful as the information it can learn from," she says. "If the most meaningful data remains locked away because it cannot be used safely, then organizations are limiting their own potential."
In her view, the companies that succeed will be those that learn how to unlock insight from sensitive data while maintaining the trust and protections that modern digital ecosystems require. "We have built this solution so others don't have to." She adds, "We are giving companies the ability to unlock value that was once trapped inside of data silos."
