Cillian Kieran (Ethyca)

Cillian Kieran (Ethyca): Launching Astralis, the AI Governance Layer

Cillian Kieran unveils Astralis, Ethyca's 'large language risk and regulatory model' acting as a control layer between enterprise data and the models businesses use, the culmination of seven years of R&D atop the company's widely adopted data governance ontology, with nearly $45M raised to date including from Capital One Ventures. He argues frontier models are becoming commoditized so proprietary data will matter most, and describes 'policy as code' as the way to stop AI agents overreaching. The conversation also covers Chinese open-source models, sovereignty versus security, Jensen Huang's defense of closed models, and consolidation in the model market.

1:22:19