Overview
Intelligence Security Laboratories (ISL) is a nonprofit research lab focused on securing advanced AI infrastructure. It designs, builds, and tests high-assurance systems to protect frontier AI models and their supporting infrastructure from sophisticated threats, including nation-state actors. Its primary initiative prototypes a secure inference data center to prevent theft, alteration, or misuse of frontier models. ISL uses a no-fail safety mindset drawn from avionics and reactor control, applying Systems-Theoretic Process Analysis for Security (STPA-Sec) to treat the entire facility as a single system and mitigate conditions that could lead to any failure. Founded by Gabriel Kulp and funded by philanthropy, the leadership includes Paul Murley and Tom Gardiner. Unlike commercial vendors, ISL aims to develop knowledge and artifacts and share them with transition partners to enable safe deployment of transformative AI, rather than operating high-security data centers for profit.
Work here?Claim your company
Simplify's take
What believers are saying
- The lab is scaling quickly, with around ten people and multiple open roles.
- Its 2026 work on GPU side channels matches urgent model-theft and verification threats.
- A recent RAND publication gives ISL a concrete technical roadmap for customers and partners.
What critics are saying
- The organization depends on philanthropy, so one funding pause can freeze hiring and prototypes.
- Its niche market is tiny; data-center operators can delay adoption until a breach occurs.
- A single severe incident during testing would destroy trust and kill transition partnerships.
What makes Intelligence Security Laboratories unique
- ISL targets secure inference data centers, not generic enterprise cybersecurity.
- Its no-fail, avionics-style process and STPA-Sec distinguish it from typical AI labs.
- RAND-derived research and philanthropic funding support public-interest work without commercial pressure.
Help us improve and share your feedback! Did you find this helpful?