AI has made its way into every corner of software development — from generating code and making decisions to powering user-facing products. But with great power comes the great potential for a spectacular failure. Join Ken Muse and Mickey Gousset as we present a series of real-world case studies showcasing some of the most cringe-worthy, harmful, and avoidable mistakes developers and organizations have made with AI.

Learn about AI-sponsored deployments gone wrong, agents that blackmail users and publish sensitive data, support chatbots going rogue, the "deadly apologist", and more. We'll dissect what went wrong and, more importantly, what should have been done differently. We will also explore some of the biggest mistakes that we see teams continue to make with AI.

Whether you're building AI-powered features or consuming AI tools in your workflow, this session will arm you with the awareness and practical strategies to keep your projects off the front page (and out of our sessions)

You will learn:

• Explore real-world case studies of AI failures and understand the root causes behind each.
• Practical guardrails and engineering practices to responsibly integrate AI into your development workflows
• Develop a framework for identifying AI risk patterns in your own projects