InfoQ talk: how to red-team GenAI systems against data poisoning and LLMjacking

A security researcher walks through adversary-emulation techniques for stress-testing GenAI apps before attackers do.