Loading…

Saturday September 12, 2026 10:30am - 11:30am CDT
The bot detection playbook defenders have relied on for years — IP blocklists, rate limits, behavioral baselines, CAPTCHA — was built for a threat that no longer exists. Modern adversaries are deploying LLM-powered agents that reason, adapt, and evolve their behavior in response to detection. For defenders, this means the threat model has fundamentally changed.   This talk, drawn from production experience building bot mitigation systems at Amazon, provides blue teamers with a practical framework for detection engineering against agentic AI attackers. The session covers: how to identify the behavioral signatures of LLM-driven agents (and why they're different from both humans and traditional bots); detection signal categories that remain robust against adaptive adversaries; pipeline architecture for high-velocity threat detection at scale; and incident response workflows when an AI-powered attacker is actively evading your controls.   Critically, this talk addresses the strategic challenge defenders face: in an adversarial ML environment, your model is always at risk of being reverse-engineered and evaded. How do you build detection systems that are robust to an adversary who can iterate as fast as you can? Attendees will leave with detection engineering patterns they can apply to bot defense, fraud prevention, and automated threat response — and a realistic understanding of where current defenses still have gaps.
Speakers
avatar for Shashwat Jain

Shashwat Jain

Sr. Software Development Engineer, Amazon
Shashwat Jain is a Senior Software Development Engineer at Amazon, where he architects and deploys AI-powered bot mitigation systems protecting Amazon's global e-commerce platforms from sophisticated automated threats. With expertise spanning real-time behavioral detection engines... Read More →
Saturday September 12, 2026 10:30am - 11:30am CDT
Swissôtel Chicago 323 E Wacker Dr, Chicago, IL 60601, USA
  Talk

Sign up or log in to save this to your schedule, view media, leave feedback and see who's attending!

Share Modal

Share this link via

Or copy link