Mindgard
Automated continuous AI red teaming from a Lancaster University spinout.
About
Mindgard is an automated AI red-teaming platform spun out of Lancaster University, treating AI security testing as its own discipline ("DAST-AI" — dynamic application security testing for AI). It runs continuous adversarial attacks against LLMs, agents, and multimodal models — jailbreaks, prompt injection, data leakage, model theft — with an attack library maintained by its own research team, which has disclosed 150+ vulnerabilities in production AI products.
Findings map to OWASP LLM Top 10 and MITRE ATLAS and plug into CI/CD. The company raised a $30M Series A in August 2026. It is testing-only by design: no quality evals, no runtime gateway, no AI inventory — it finds attacks, and enforcement lives elsewhere.
What it can do
- Continuous automated red teaming (LLM, agent, multimodal)
- Research-maintained attack library (150+ disclosed vulns)
- CI/CD integration with pass/fail thresholds
- OWASP LLM Top 10 and MITRE ATLAS mapping
- Remediation advice and quantified risk reports
Setup
- Book a demo at mindgard.ai.
- Provide an inference endpoint for your model or agent.
- Select attack scenarios or run the full library.
- Review quantified risk reports mapped to OWASP/MITRE ATLAS.
- Integrate with CI/CD for continuous re-testing on every change.
Prerequisites
Commercial account; an inference endpoint for the system under test.
Use cases
Works with
Trust score — 6/10
Scored 6/10 (metrics tier, metrics-v1). Evidence: Lancaster University spinout; $30M Series A led by Album VC reported Aug 2026; Claims 150+ vulnerabilities disclosed in production AI products; Testing-only by design: no runtime gateway, no quality evals, no AI inventory; Commercial SaaS, demo-gated; independent verification of attack library not available. License: Proprietary SaaS.
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 6.0/10
- Maintenance · recency, releases, issue load
- 7.0/10
- Openness · license + docs
- 4.0/10
- Manual review · sanity read + red-flag check
- 7.0/10
- · Lancaster University spinout; $30M Series A led by Album VC reported Aug 2026
- · Claims 150+ vulnerabilities disclosed in production AI products
- · Testing-only by design: no runtime gateway, no quality evals, no AI inventory
- · Commercial SaaS, demo-gated; independent verification of attack library not available
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.