Mindgard is an AI security platform with an offensive focus. It discovers AI assets and attack surfaces, including shadow AI, runs adaptive AI red-teaming to find exploitable weaknesses in models, agents and applications, provides remediation guidance, retests fixes and offers runtime protection. The platform is aimed at security, engineering, GRC and audit teams, mainly in enterprises, and connects through APIs, CI/CD pipelines and Burp Suite. The company grew out of AI security research at Lancaster University and reports SOC 2 Type 2 compliance. It is not open source.
According to its website and press coverage, Mindgard has locations in London and Boston, and it is described in the press as a UK spin-out. The UK is outside the EU and the EEA. The website does not state which hosting regions or deployment models (SaaS, on-premises) are available. European organisations should settle this with the vendor in advance, because red-teaming can involve test prompts, system configurations and results.