// TECHCRUNCH — INTELLIGENZA ARTIFICIALE
Get all your questions answered at TechCrunch Disrupt 2026: The full breakout session agenda revealed
TechCrunch Disrupt 2026 lands at Moscone West in San Francisco on October 13-15, bringing together 10,000+ founders, investors, operators, and tech leaders for three days of ideas, connections, and conversations about what’s coming next. Register now, save up to $100 off your pass, and get a second at 50% off.
Some of the most practical conversations won’t happen on any of the industry stages. Disrupt’s breakout sessions are built to go deeper and give you the mic to ask your burning questions to the leaders. Each 50-minute session combines expert insight with audience Q&A, giving you more room to unpack a problem, challenge assumptions, and leave with ideas you can actually put to work. The room capacity is limited, and attendance is first come, first served.
This year’s lineup gets into the questions founders and operators are dealing with right now: How do you build with AI agents? Where will value accrue as AI shifts from training to inference? What does physical AI really mean? How do you run a better fundraiser? And how does AI change the way you grow companies and teams?
Get into the room, get your questions answered, and make the connections that’ll spark your next move in the ecosystem. Save up to $100 on your pass and 50% on a second. Bring a group of four-plus to save more and get more insights. If you’ve recently been laid off, get a $75 Expo+ Pass before they run out.
Get to know the sessions and tech leaders shaping the conversation on the Disrupt agenda.
Susan Schofer, Partner and Chief Science Officer, SOSV HAX; and Duncan Turner, General Partner and Managing Director, SOSV HAX
AI is moving beyond software and into machines that need to perceive, reason, and act in unpredictable environments. Schofer and Turner break down what actually qualifies as physical AI, why building it requires a different playbook, and what investors are looking for in companies trying to define the category.
Steve Androulakis and Sachin Malhotra, Members of Technical Staff, Anthropic
What does an AI-agent workflow look like when the people using it build AI for a living? Anthropic’s Androulakis and Malhotra show how they run multiple Claude agents in parallel, delegate real engineering work, review output, recover when agents get things wrong, and make agents part of their everyday jobs.
As AI takes on more consequential work, accuracy alone isn’t enough. Drawing on experience spanning misinformation, online safety, and work with foundation-model labs, Goldfarb explores how expert human judgment can become scalable evaluations — and what it will take to build AI systems people can actually rely on.