TechCrunch Disrupt 2026, running October 13–15 in San Francisco, will feature a dedicated AI Stage presented by Google for Startups. The agenda focuses on practical challenges founders face when building and scaling AI-native businesses: pricing AI products as models commoditize, rebuilding security from the infrastructure up for agentic systems, and developing go-to-market strategies that did not exist two years ago.

Sessions include Anthropic‘s Head of Applied AI Cat de Jong discussing what enterprise deployments actually look like after launch—where they succeed, where they stall, and what separates real value extraction from prolonged pilots. OpenAI‘s Head of Productivity Tara Seshan will address how AI collapsed the traditional go-to-market stack and created the new discipline of GTM engineering, now among the fastest-growing roles in tech. Databricks co-founder Arsalan Tavakoli will cover enterprise AI security requirements in 2026, including observability, governance, and the architecture that earns trust for autonomous decisions inside sensitive systems.

Okta’s Ric Smith will talk about agent security at the infrastructure level, arguing that application-level permission models are fundamentally flawed for agentic AI. Decart and Luma AI’s CEOs discuss real-time visual AI inference and physical reasoning. A panel with CEOs from Glean and Monte Carlo, along with venture partners, tackles whether the SaaS playbook is dead or evolving—covering sustainable AI pricing, defensible moats under model commoditization, and how to make SaaS work in the AI era. Clay’s CEO Kareem Amin will present on the GTM engineering role. AWS’s Chet Kapoor, Katie Moussouris, and Wendy Nather will deliver an infrastructure-level view of enterprise AI security.

The conference expects 10,000+ startup, tech, and VC leaders across all stages and the exhibition floor. Early pricing with up to $200 savings is ending soon.