Navigating the Challenges of Regulated AI: Insights from Ketryx CEO Erez Kaminski
Deploying regulated AI is no small feat, as highlighted by Erez Kaminski, CEO of Ketryx and former AI strategist for Amgen. With the FDA’s recent rollout of its internal AI tool, Elsa, the complexities of regulatory compliance are becoming increasingly apparent. Kaminski emphasizes that the challenges go beyond just the sheer volume of regulatory documents; they also involve the architecture of AI systems themselves. He advocates for a hybrid approach that combines the strengths of large language models with the structured logic of symbolic AI, enabling a more efficient and effective review process.
The intricacies of regulatory environments, particularly in healthcare, require a nuanced understanding of both technology and compliance. Kaminski points out that the development of medical devices and pharmaceuticals involves a branching path of documentation that far exceeds the complexity of typical AI applications. As the FDA grapples with staffing challenges and the overwhelming amount of information, the need for innovative AI solutions becomes even more critical. By integrating neuro-symbolic frameworks, the potential for AI to enhance human capabilities in regulatory science is immense.
Looking ahead, the goal is to achieve “accountable autonomy” in AI systems, allowing them to operate independently while ensuring traceability and safety. As Kaminski notes, the floodgates are open for AI to transform the regulatory landscape, but it requires a concerted effort to harness its potential effectively. How can we ensure that these advancements lead to better healthcare outcomes while maintaining rigorous compliance standards?
Original source: https://www.rdworldonline.com/as-fda-moves-builds-out-elsa-this-ai-compliance-ceo-underscores-that-need-for-a-hybrid-ai-approach/