AI agents and chatbots increasingly run Python code interpreters that can be abused by attackers. This talk shows how prompt-to-RCE, credentials abuse, and C2 payloads work in these sandboxes, including a real AWS AgentCore breakout, and how to harden against code interpreter exploits.
AI agents and chatbots increasingly run Python code interpreters that can be abused by attackers. This talk shows how prompt-to-RCE, credentials abuse, and C2 payloads work in these sandboxes, including a real AWS AgentCore breakout, and how to harden against code interpreter exploits.
Guest appearance on a the Cloud Security Podcast, discussing AWS Security Guardrails, CI/CD, Observability, Serverless, and more.
In this talk, I cover what sole security practitioners need to build out their own 'lean but mean' cloud security toolkit and provide demos of the suggested tools.
Azure Guardrails, published by Kinnaird in July 2021, helps you deploy hundreds of security guardrails in Azure within minutes. He discusses the tool in this talk and provides a demo.
SHIFTing Left by Creating Guardrails using AWS Service Control Policies and Making Least Privilege Easy for Developers
Policy Sentry, published by Kinnaird in late 2019, helps you write least privilege AWS IAM Policies. He discusses the tool in this talk and provides a demo.
Guest appearance on a Podcast, discussing open source IAM security tooling
When you’re operating in a cloud environment, access expands, responsibilities change, control shifts, and the speed of provisioning resources and applications increases—significantly affecting all aspects of IT security. Security must keep up with …