Mindgard is a continuous, automated AI red teaming platform for testing LLMs and AI agents against adversarial attacks, aligned to the MITRE ATLAS framework, spun out of AI security research at Lancaster University and formally founded in 2022 in London. It has raised approximately $8 million in funding and holds SOC 2 Type II certification. The platform focuses on ongoing offensive security testing rather than inline runtime blocking. Pricing is enterprise quote-based.
ai-securityred-teamingmitre-atlas
Overview
- Founded
- 2022
- Pricing tier
- High
- Startup-friendly
- No
- Enterprise-ready
- Yes
Agent security
- Prompt injection detection
- Yes
- Red teaming
- Yes
- Runtime protection
- No
- Model scanning
- No
- Open-source
- No
- Self-hostable
- No
- Free tier
- No