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

Founders building AI products can no longer ignore safety and security as afterthoughts; they are now key differentiators for enterprise adoption.

TechCrunch Disrupt 2026 will feature five AI safety sessions aimed at startup founders. These sessions will cover critical aspects of developing and deploying AI systems responsibly. Topics include enterprise AI deployment challenges with Anthropic's Head of Applied AI, agent security risks with Okta and NanoCo, and the complex security and governance needs for AI in enterprise cloud environments, featuring experts from AWS, Luta Security, and cybersecurity veteran Wendy Nather. Additional sessions will address building AI systems where failure is not an option, drawing from defense and autonomous vehicle expertise with Shield AI, General Motors, and Waabi, and exploring the data and infrastructure challenges hindering robotics' progress, with insights from NVIDIA. The event, scheduled for October 13-15 in San Francisco, aims to equip founders with knowledge on building trustworthy AI.

The focus on AI safety at TechCrunch Disrupt underscores a maturing market where practical deployment and user trust are becoming paramount. Sessions like Anthropic's on enterprise adoption and Okta's on agent security highlight the shift from theoretical capabilities to real-world integration challenges. Companies building AI for business customers must now proactively address security vulnerabilities and governance frameworks, as highlighted by the AWS and Luta Security panel. This indicates a growing demand for demonstrable safety and reliability, moving beyond impressive demos to robust, secure, and compliant AI solutions.

For AI companies, especially those targeting critical infrastructure or physical applications, the emphasis on failure mitigation and validation is significant. The discussions around robotics and autonomous systems, featuring leaders from Shield AI and NVIDIA, point to the need for rigorous testing, safety cultures, and addressing data limitations to achieve mainstream adoption. Founders must consider these aspects early in their development lifecycle, as regulatory hurdles and customer skepticism will increase for AI systems operating in unpredictable environments. The event signals that the next wave of AI innovation will be judged not just on capability, but on its inherent safety and trustworthiness.