We are excited to announce that Matchstick Ventures Fund III has completed an investment in Liminal. Liminal delivers horizontal security for all generative AI. With Liminal, CIOs and CISOs take back complete control over AI data privacy, security, and sovereignty.

Generative AI is widely seen as the greatest leap in productivity in a generation. Even in its early stages, gen AI is prolific. Mid-market and large enterprises have shown a strong appetite for directly interacting with large language models (e.g. GPT4), building new AI enabled applications, and consuming the myriad gen AI enabled applications now available off-the-shelf.

CIOs and CISOs within regulated industries are grappling with increased pressure to develop and execute a strategy around AI while maintaining a strong security posture. Currently, most are imposing a strict, no-use policy around generative AI, simply because they lack the technology to monitor, observe, and enforce policies that can govern gen AI usage.

Liminal empowers regulated enterprises to securely deploy and use generative AI across all use cases. With Liminal, CIOs and CISOs have complete control over the data submitted to large language models (LLMs), ensuring they are protected from regulatory compliance risk, data security risk, and reputational risk.

For direct chat interactions with LLMs, Liminal provides a secure experience for organizations to govern access to models they have licensed, and enforce granular security controls over what data is able to be shared with each model. Liminal’s engine detects PII, PHI, PCI, IP, and other sensitive data types and accurately redacts or masks these terms from a prompt prior to its submission. Liminal’s masking capabilities are industry leading, ensuring critical context is maintained while adhering to corporate, state, and federal data privacy guidelines.

Liminal also provides an easy to deploy and manage desktop application. This empowers mid-market and enterprise CISOs to similarly enforce granular security policies within the generative AI components of off-the-shelf software (e.g. Office 365 Copilot), while preserving the intended user experience.

These same organizations can govern usage of applications they build that contain generative AI. By integrating just 4-lines of code, Liminal secures all interactions with the gen AI component of their apps.

Liminal is the only platform to provide horizontal security across all generative AI. With their world-class admin experience, observation tools, and fine-grain policy controls, they are rapidly emerging as the security solution for CIOs and CISOs in regulated enterprises. Whether organizations are just beginning their AI journey or have a comprehensive AI strategy, Liminal is able to capably meet their demands, now and in the future.

We met Steven Walchek, the CEO of Liminal, through the Colorado startup ecosystem, and have worked closely with him and his team, as well as High Alpha who are co-investors with us on this investment.

Within less than 4 months, Liminal has achieved commitments from a Fortune 150 life sciences company, one of the largest healthcare systems in the US, and a number of mid-market banks. We are impressed by their early success, and are excited to enable their future growth.