Agentic Security - BSides NYC Week

​Agentic Security NYC is a practitioner-only gathering for security engineers and leaders actually running AI agents in production. Not a conference, and not a place to pitch a product.

​We ran the first one in August. The second is co-hosted with Datadog at their New York office on Friday, October 16, the night before BSidesNYC.

​Program

• ​6:00 Doors, food and networking

• ​6:30 Welcome from Agentic Security NYC and Datadog

• ​6:40 Offensive AI security talk: how agents actually get attacked [speaker and title TBA]

• ​7:10 Datadog defensive security talk: detecting and responding in production [speaker and title TBA]

• ​7:35 Moderated practitioner discussion: what actually broke, who has the authority to say no to an agent going to production, whether you could prove what an agent did last week, and where your line on autonomy really sits

• ​8:30 Networking

​Rules of the room → Chatham House for the discussion: share the ideas afterward, not who said what → No vendor pitches, from us or anyone else → Practitioners only

​Photography: a photographer will capture the room. Crowd shots are wide or from behind, speakers are photographed with consent, and anyone who'd rather not be photographed can pick up an opt-out sticker at check-in. The discussion is never recorded.

​Getting in: the address and check-in instructions go to approved guests. Register with the name on your photo ID.

​If you're building or defending AI agents in production (security engineering, AppSec, ProdSec, InfraSec, AI governance), request an invite. Space is limited.