Lean SuperIntelligence Launches Self-Improving Cybersecurity Models, Secures Initial Funding
Lean SuperIntelligence (LSI) has launched two new, self-improving cybersecurity models designed for on-premise deployment. The company also announced the closing of its initial financing round.
Palo Alto, CA – Lean SuperIntelligence (LSI) has introduced two new AI models, LSI-Offense-1 and LSI-Defense-1, aimed at enhancing enterprise cybersecurity through coordinated offense and defense capabilities. These models are designed to operate locally on a customer's infrastructure, eliminating the need to transfer sensitive data to cloud-based systems.
The LSI-Offense-1 LLM identifies and exploits vulnerabilities, while the LSI-Defense-1 LLM detects and responds to threats. The company states that the models continuously improve each other; the offense model discovers new vulnerabilities, and the defense model learns from these findings, leading to faster threat response times.
LSI claims its specialized models outperform larger, general-purpose frontier models in vulnerability discovery and threat detection, operating at a significantly lower cost. The company's founding team includes individuals with prior experience at prominent technology firms such as OpenAI, CrowdStrike, and SentinelOne.
In addition to the model launches, LSI announced the closing of its initial financing round, supported by several investors and angel investors. This funding is intended to support the company's ongoing development and expansion efforts.
The new models are currently available to a select group of design partners, including cybersecurity vendors and Fortune 500 enterprises, for air-gapped, on-premise deployments.