The event, held in partnership with Trust Valley, brought together thought leaders to discuss the critical intersection of Generative AI (GenAI), safety, and compliance. The engaged audience and inquisitive questions truly highlighted the urgent need for robust frameworks in this rapidly evolving space.
Key Takeaways from Industry Experts
The session featured excellent speakers who provided valuable perspectives:
Pascal Marmier of Unlimitrust and Trust Valley opened the session, setting the stage for a deep dive into AI trust and safety, which is central to the mission of both organizations.
Prof. Andrei Kucharavy delivered expert insight into the inherent challenges of securing AI, particularly GenAI applications. He underscored the technical complexities that must be addressed to ensure AI systems operate reliably and securely.
Terry Vogelsang offered a practitioner’s view, emphasizing the necessity of applying established security-by-design principles to GenAI. He highlighted the difficulty in properly testing GenAI applications due to their probabilistic (as opposed to deterministic) output, as shown in his slide:
“In a traditional computer program: A + B = C”
“In an LLM-based Application: A + B = C${70%}$ or D${66%}$ or …”
He also stressed the meticulous care required when implementing Agentic AI applications, detailing the need for controls on data access and independent action capabilities.
Gilles Burnier concluded the presentations from a business leadership perspective, focusing on the essential task of balancing risk and innovation while stressing the overriding need for compliance. His presentation on “Operational and Reputational Risks” clearly illustrated the pitfalls of neglecting these controls.
Networking and Collaborative Spirit
Beyond the insightful presentations, the event fostered a vibrant networking session where ComplianceRT’s Sam had many engaging conversations with attendees. It was a fantastic opportunity to connect with fellow professionals, including Gaetan Stein, Cristina Pereiro, Evelina Georgieva, and the speakers themselves, Terry Vogelsang and Gilles Burnier. These discussions further underscored the shared commitment within the community to collectively tackle the challenges and opportunities presented by AI.
The Growing Need for AI Compliance and Trust
The central theme resonated strongly with ComplianceRT’s mission: as AI, and GenAI in particular, becomes integrated into core business processes, the operational and reputational risks escalate. Issues such as hallucinations and misinformation, brand misalignment, and the difficulty in securing predictable outcomes demand a systematic approach to AI governance.
This is where compliance becomes not just a regulatory obligation, but a strategic imperative. The discussions provided an excellent segue to the vital work we do at ComplianceRT.com.
ComplianceRT specializes in AI and Cybersecurity compliance, helping organizations navigate the complexities of:
- The EU AI Act: Preparing for and achieving compliance with the world’s first comprehensive AI regulation.
- ISO/IEC 42001 (AI Management System): Establishing a robust framework for managing AI responsibly and ethically.
A proactive approach to compliance—applying security by design, rigorous validation and stress-testing (as championed by Surelio.ai), and establishing clear governance—is the only way to effectively balance risk and innovation in the AI era.
We look forward to collaborating with leaders and practitioners to build a trustworthy and safe AI future.
- We can help you become FADP compliant!
Expert Guidance, Affordable Solutions, and a Seamless Path to Compliance
