Speaking

I talk about security for AI systems: prompt injection, agent security, and red teaming LLM applications. Most of it comes from building Promptfoo and now working on cyber products at OpenAI.

Public sessions

  1. PanelSan Francisco

    The AI Security Panel: Unfiltered

    Hosted by Snyk the day before the AI Security Summit.

    (opens in a new tab)
  2. PanelNew York

    The Compounding Intelligence Race

    At ScaleUp:AI, Insight Partners' AI conference.

    (opens in a new tab)
  3. PanelLas Vegas

    Skill Issue: An Opening Panel for the OWASP Top 10 for Agentic Skills Initiative

    The public launch of the OWASP Agentic Skills Top 10, during Black Hat USA.

    (opens in a new tab)
  4. Arsenal LabToronto

    SecTor

    Live demos of Promptfoo's open-source AI red teaming, with the Promptfoo team.

    (opens in a new tab)
  5. Arsenal LabLas Vegas

    Black Hat USA

    Live demos of Promptfoo's open-source AI red teaming, with the Promptfoo team.

    (opens in a new tab)

Closed-door sessions

I also run closed-door sessions with security and engineering teams on behalf of OpenAI, which aren't listed here. If you're organizing an event or working on the same problems, get in touch.