top of page


GuardAI provides a trustworthy AI tool for large language model developers and users to automatically detect and defend vulnerabilities, such as bias, privacy breach, and harmful messages,  to ensure AI is safe and reliable.



  • Prioritizes trust and ethical standards in the field of the large language models.

  • Blends attack strategies, robust defense mechanisms, and comprehensive risk assessment framework,  adapting to the evolving LLM landscape.

  • Boasts cutting-edge features like advanced role-playing mechanics and automatic jailbreak identification, ensuring unparalleled robustness and reliability.


GuardAI is more than a tool. It is our commitment to a future where AI thrives responsibly and reliably.

Meet the Tech-Founder

  • LinkedIn

Haohan Wang, the mastermind behind GuardAI technology, is an assistant professor in the School of Information Sciences at the University of Illinois Urbana-Champaign. He is also affiliated with the Computer Science Department and the National Center for Supercomputing Applications at UIUC. His research focuses on developing trustworthy machine-learning methods and applying these technologies to complex applications. With a decade of experience, his expertise spans a wide spectrum, from rigorous statistics to powerful deep learning and large language models. He has extensively studied the trustworthiness challenges associated with these technologies. Wang has received several recognitions from various institutions and agencies. He earned his Ph.D. in computer science through the Language Technologies Institute at Carnegie Mellon University.

The Genesis of GuardAI: A Convergence of Vision, Expertise, & Innovation

Our startup's journey began with a chance meeting at CMU's Innovation Summit in 2015, where Jenny sat on a panel and connected with her future business partner, Professor Wang. ​With over two decades of experience in managing financial model risk, Jenny was well-versed in the complexities of regulatory guidance. However, her confidence in model risk management was truly challenged by the advent of large language models. Inspired by these challenges, Jenny penned a LinkedIn article titled "You Cannot Manage AI without an AI Tool," following the European Parliament's adoption of the Artificial Intelligence Act on June 14, 2023. This article marked her proactive search for cutting-edge technological solutions. ​ On August 13, 2023, a breakthrough came. Jenny noticed a post by Professor Wang detailing his extensive research with scholars, studying 673 papers on trustworthy machine learning over the last 20 months. Immediately, she reached out to him, sensing a unique opportunity. It was a moment that led to the right technology meeting the right business need. Our shared passion for building trust in AI and blending emerging business needs with advanced technology culminated in the birth of our startup. We embarked on a mission to create scalable, trust-infused AI solutions, harnessing our combined expertise to make a mark in the AI landscape.

bottom of page