Heron
Heron
Connecting cyber talent with AI leaders and opportunities
The Heron Program connects cybersecurity professionals to the growing field of AI security. For 6 weeks, participants will spend an evening exploring frontier AI security through hands-on projects and Q&A sessions with global experts. Select participants will be invited to an AI Security conference in Berkeley, California.
Why ai security?
As AI systems become more powerful, the stakes of a security breach skyrocket. Most labs are currently far short of the security level needed to defend against motivated state actors, and governments have yet to develop adequate oversight.
A RAND report from May 2024, led by Sella Nevo and Dan Lahav, addresses the current security state of frontier AI labs. The report identifies significant gaps in the security of model weights relative to the current capabilities and threat models, and was one of the inspirations for the creation of the Heron Program.
A 6-week initiative that brings together top-tier cybersecurity experts to explore the emerging field of frontier AI security. Through exclusive in-person weekly meetings in Tel Aviv, featuring Q&A sessions from industry leaders and hands-on projects, participants will dive into crucial topics like model weight protection, edge deployment security, and AI-specific attack vectors, with the dual aim of advancing the field and fostering the next generation of frontier AI security specialists.