Q-Cyber’s Kevin Zwaan to reveal groundbreaking “Claude Hack” at Exclusive Roundtable in The Hague
Q-Cyber, a leading Dutch cybersecurity and penetration testing consultancy, today announced that its ethical hacker, Kevin Zwaan, will be the featured speaker at the exclusive roundtable hosted by Anove in cooperation with the CCI France Pays-Bas.
From Guardrails to Malware Factory
The event, titled “From Guardrails to Malware Factory: Lessons from the Claude Hack,” will take place on June 3rd in The Hague.
The roundtable will focus on the urgent and evolving landscape of Artificial Intelligence security, highlighted by Zwaan’s recent, internationally recognized demonstration. Zwaan successfully bypassed the safety guardrails of Anthropic’s Claude AI in just eight hours, transforming the model into a potential malware generator. Together with Professor Yuri Bobbert, he is now presenting this case to the specialist and audiences involved in AI, Cybersecurity and Governance.
Radicalization as a Service
Unlike traditional cyberattacks that uses bugs (living of the land), Zwaan utilized psychological manipulation—a technique he has named “Radicalization as a Service.”
By exploiting the AI’s programmed motivations, he demonstrated that even the most advanced Large Language Models (LLMs) can be convinced to abandon their security protocols.
“The guardrails of an LLM are hardcoded, making their motivations fixed and predictable,” said Pierre Kleine Schaars, COO at Q-Cyber. “By understanding these motivations, we can manipulate the model into trusting us over its own safety guidelines. This isn’t just a theoretical vulnerability; it’s a scalable method that could be used to automate phishing, generate malware, or spread disinformation on a massive scale. We are thrilled that Anove recognizes the urgency of this issue and is providing a platform to share this.”
The roundtable will bring together CISOs, security leaders, and AI experts to discuss the implications of this new threat. Zwaan will break down the mechanics of the attack and explain why AI security demands a fundamental shift in approach. He will be joined by Jean-Hugues Migeon, CEO of Anove and an expert in AI governance, who will discuss how frameworks like Anove’s AI Management System (AIMS) combined with Q-Cyber’s advanced penetration testing can secure corporate AI portfolios.
“This invitation from Anove is a significant validation of the cutting-edge research we conduct at Q-Cyber,” a Q-Cyber spokesperson added. “As organizations rapidly adopt AI, they must understand that traditional security measures are no longer sufficient. We are committed to helping businesses navigate these complex new risks.”
Attendance at the roundtable is limited and subject to confirmation based on professional relevance. Interested security and IT leaders can request an invitation via the Anove website. https://www.cfci.nl/actualites/n/news/roundtable-anove.html