Guardrail Technologies
A Park City, Utah startup selling a provider-agnostic "Trust Layer" that controls what data generative and agentic AI models can see, store and reveal.
Visit Website ↗ + Add to Compare Claim This CompanyOverview
Guardrail Technologies sells a control and governance layer, which it calls the Trust Layer, that sits between users and underlying generative or agentic AI models. It applies context-aware masking of sensitive data, prompt-level policy rules, role-based access control and audit logging, and is built to work across models from multiple providers so enterprises can switch or mix AI vendors without rebuilding their safety and privacy controls each time.
Based in Park City, Utah and founded in 2020, the company restructured its leadership in 2025, naming Todd (T.J.) Marlin as CEO, with Filter Fund founder Richard Sussman becoming Executive Chairman after leading its investment. In October 2026 it announced close to $3 million raised across two seed rounds led by Filter Fund Venture Capital, with participation from strategic angel investors, earmarked for expansion into Asia (building on existing Australian operations) and Europe.
Its stated differentiator is being provider-agnostic rather than tied to one model vendor’s own guardrails, aimed at large enterprises and regulated industries such as financial services, insurance and healthcare, though the broader AI-governance space already includes better-funded players pursuing similar positioning.
Innovation Matrix Assessment
Recently restructured leadership (new CEO in 2025) and closed a second seed round in October 2026; moving, but with a long gap between 2020 founding and visible commercial traction.
A provider-agnostic trust/control layer addresses a real multi-vendor AI governance pain point for enterprises that use more than one model provider.
Under $3 million raised in total across two seed rounds, and no named enterprise customers were found in this research.
Provider-agnostic AI governance is a reasonable differentiator against single-vendor guardrail tools, but the AI-governance space (Guardrails AI, WitnessAI, Lakera and others) is already crowded and generally better funded.
No independent evidence of the Trust Layer's real-world effectiveness was found; available descriptions are from company and directory materials.
Cross-provider AI governance should stay relevant as enterprises adopt multiple model vendors, though this company's small scale makes its own staying power uncertain.
Why CISOs Should Care
Offers a single governance layer for generative and agentic AI usage that doesn't need to be rebuilt each time the organization adds or swaps a model provider.
What Makes It Different
Explicitly provider-agnostic positioning, aimed at letting enterprises mix AI vendors while keeping one consistent set of privacy and safety controls.
The Matrix Verdict
47/100 — EMERGING / UNRANKED
An Emerging/Unranked company: a sensible pitch and real product, but currently too small and unproven in the market to rank higher.
Editorial Note: Claims vs. Verified Findings
Product capability descriptions come from the company's own site and third-party directories; no independent verification of effectiveness or named customers was found.
Sources
Alternatives to Guardrail Technologies
Adaptive Security
AI-driven platform that simulates deepfake, voice, and multichannel social-engineering attacks to train and test organizations against next-generation phishing.
Quilr
CISO ReviewedEarly-stage agentic AI security startup building a 'Service-as-Software' platform to guard against human-related breaches and secure AI agent…
Zenity
Governance and security platform for AI agents and low-code/no-code development, securing agent identity, permissions and behavior across the…
Tenzai
An agentic AI penetration testing startup building autonomous 'AI hackers' to find and validate exploitable vulnerabilities at a…
Gray Swan AI
Carnegie Mellon-linked startup building automated red-teaming and runtime guardrails that frontier AI labs use to stress-test their models…
Alice (formerly ActiveFence)
Israeli AI security company (rebranded from ActiveFence in January 2026) offering a lifecycle platform to test, guard, and…