Five AI safety sessions every founder should have on their TechCrunch Disrupt 2026 agenda

TechCrunch Disrupt 2026 will feature five sessions focused on AI safety across the AI Stage and Real World AI Stage, addressing risks from enterprise deployments to autonomous systems and robots. Speakers include leaders from Anthropic, Okta, AWS, Shield AI, GM, Waabi, Nvidia and others who will discuss security, governance, testing, and data challenges as AI moves into production and the physical world.

By AI Newsroom· Reviewed by Pranav, Founder & Editor-in-ChiefPublished about 1 hour agoUpdated about 1 hour ago0 views

Five AI safety sessions every founder should have on their TechCrunch Disrupt 2026 agenda

Why It Matters

As AI shifts from demonstrations to real-world use in enterprises, vehicles and robots, founders must confront safety, security and data limitations that can block deployment; these sessions offer practical insights from vendors and operators actively facing those challenges. Understanding these issues is crucial for builders hoping to move AI products from pilots into trusted production systems.

Key Facts

Event: TechCrunch Disrupt 2026
Number of featured AI safety sessions: Five
Stages: AI Stage and Real World AI Stage
Speakers/Organizations mentioned: Anthropic (Cat de Jong), Okta (Ric Smith), NanoCo (Gavriel Cohen), AWS (Rudy Mitra), Luta Security (Katie Moussouris), Wendy Nather, Shield AI (Nathan Michael), General Motors (Mikell Taylor), Waabi (Raquel Urtasun), Nvidia (Les Karpas)
Topics covered: Enterprise deployments of Claude, agent security, cloud security for AI, safety for physical autonomous systems, robotics data challenges and foundation models.

TechCrunch Disrupt 2026 will dedicate five sessions to AI safety and security across its AI and Real World AI stages, assembling practitioners and vendors who are wrestling with putting AI into production and the physical world. The programming targets founders and product teams confronting questions about enterprise readiness, agent permissions, infrastructure risk, and safety for autonomous systems. Anthropic’s Head of Applied AI Cat de Jong will discuss patterns she observes when companies move Claude from pilot projects into critical workflows, highlighting factors that help some deployments deliver measurable business value while others stall in pilot mode. That session aims to give founders selling enterprise AI a closer view of deployment barriers and success factors. A pair of sessions will examine security at different layers. Okta President of Products and Technology Ric Smith and NanoCo CEO Gavriel Cohen will address agent-level security, including access and architectural concerns when agents can take actions on behalf of users. Separately, AWS VP of Security Services Rudy Mitra, Luta Security CEO Katie Moussouris, and cybersecurity expert Wendy Nather will look at governance, observability, and infrastructure requirements as AI assumes more autonomous roles inside enterprises. On the Real World AI Stage, leaders from Shield AI, General Motors, and Waabi will talk about designing and validating systems where failures have physical consequences, covering safety culture, testing, regulatory considerations, and earning operational trust. Nvidia’s Les Karpas will examine the data and simulation gaps that limit robotics today and explore how pipelines, simulators, and foundation models could help robots scale toward broader, reliable deployment. Together the sessions frame a central question for AI founders: will people and organizations be able to trust what you build? The agenda is part of a larger Disrupt program; attendees who register before September 25 can save up to $200 on passes, and a second pass is available at 50% off.