Mindgard's DAST-AI platform is recognised for delivering adversarial testing and runtime vulnerability scanning
LONDON, Jan. 8, 2025 /PRNewswire-PRWeb/ -- Mindgard, a leader in automated AI red teaming and security testing, has been recognized in the Open Web Application Security Project (OWASP) LLM and Generative AI Security Solutions Landscape Guide 2025. The guide highlights knowledge and solutions that help organisations address the growing security risks tied to Large Language Models (LLMs) and Generative AI.
Mindgard delivers the first and only Dynamic Application Security Testing for AI (DAST-AI) solution specifically designed to detect and remediate AI-specific vulnerabilities that only manifest during runtime. The solution reduces testing times from months to minutes while continuously monitoring and mitigating risks across the entire AI lifecycle. Once deployed, Mindgard's DAST-AI provides ongoing protection as new threats emerge. Additionally, it validates the effectiveness of guardrail and WAF solutions. Seamless integration with existing CI/CD pipelines offers actionable insights and compliance-ready reporting aligned with MITRE and OWASP frameworks, making it both practical and scalable for security teams.
OWASP's 2025 Top 10 Risks for LLMs underscores the urgent need for specialized AI security solutions. It identifies significant risks such as the rise of System Prompt Leakage, which was responsible for over 30 data breaches in 2024. Other trends include resource mismanagement, which added 15% to LLM operational costs last year, and a 40% increase in attacks targeting embedding pipelines like Retrieval-Augmented Generation (RAG). Expanded risks, such as Unbounded Consumption (linked to operational cost surges), excessive autonomy in AI systems, and embedding vulnerabilities, highlight the need for proactive solutions like Mindgard's.
Dr. Peter Garraghan, CEO and co-founder of Mindgard, said: "OWASP's guide is immensely helpful for organisations seeking to understand and navigate LLMs and generative AI security solutions. Security can no longer be an afterthought when these systems are handling sensitive data and critical business operations. Continuous testing, like Mindgard's DAST-AI, is essential to uncover vulnerabilities before they can be exploited. Our platform gives organisations the tools to efficiently and effectively secure their AI systems at scale."
Mindgard's inclusion in the OWASP guide reinforces its position as a pioneer in AI security. Founded in a leading UK university lab and backed by over a decade of cutting-edge research, Mindgard's solutions are developed by a PhD-led team specialising in thousands of AI attack scenarios. Its platform provides unparalleled model coverage across LLMs, image, audio, and multimodal systems, ensuring comprehensive protection against the evolving threat landscape.
For more information on how Mindgard is securing LLM-based applications, visit www.mindgard.com
Media Contact
Joe Krasean, Mindgard, 49 17683016703, [email protected], https://mindgard.ai/
SOURCE Mindgard
Share this article