Securing AI Systems

What topics or questions you would like see addressed in the meeting discussions? Please provide a maximum of five. Note that "popular" topics are likely already considered.

The following prompts may be helpful:

Technical gaps in our ability to secure AI systems.
Security evaluation frameworks to measure the robustness, reliability, and trustworthiness of AI-enabled systems.
AI agents identity –how to establish and prove.
Boundaries between content and control in AI systems.
Least privilege: how to define, enforce, and measured.

*This question is required.

Your name: *This question is required.

Email This question requires a valid email address.

Organization/role

Disclaimer: This form information is being collected via Alchemer. Alchemer and Widgix, LLC are not affiliated in any way with, or endorsed by, the National Academies of Sciences, Engineering, and Medicine, and your submission via the Alchemer website is subject to Alchemer's privacy policy and terms of use. Any data you provide to Swoogo that is shared back with the National Academies of Sciences, Engineering, and Medicine will be treated in accordance with the National Academy of Sciences' Privacy Statement as if the data had been provided by you to the National Academies.