Would you trust an AI agent with access to your company’s systems? Would you put your employees in self-driving cars? Would you deploy robots that need to make sense of the unpredictable physical world?
For AI founders, these are no longer hypothetical questions. As AI moves from demonstrations to businesses, vehicles, robots, and autonomous agents, safety and security will become part of the product.
At TechCrunch Disrupt 2026, five sessions across the AI and Real World AI stages brought these challenges to the forefront. Here, founders can dig into what it takes to build artificial intelligence (AI) that people and businesses actually trust and use, from securing agents to putting enterprise AI into production.
Want to build the next generation of AI? Secure your pass to Disrupt and add these 5 AI safety sessions to your agenda. Save up to $200 on your pass by September 25th at 11:59pm PT. Get your second pass at 50% off.
1. What Anthropic sees when companies implement Claude in the real world
Some companies are getting measurable value from AI. Some are still running their pilots 18 months later. why?
A close-up look at Cat de Jong, Head of Humanity at Applied AI. She works directly with companies that are deploying Claude into their critical workflows to see where their deployments succeed, where they stall, and what separates the two.
In “What Humanity Sees When Companies Put Claude into Practice,” de Jong takes a look at the AI stages and explores what happens when companies move from AI experimentation to production. For founders looking to sell AI to enterprises, her experience provides first-hand insight into why some deployments gain real value while others remain in pilot mode.
Want to know the difference between an AI pilot and an implementation that delivers real business value? Hear first-hand what Anthropic sees in action. Register with Disrupt now and save up to $200 through September 25th.
2. Agent security issues that no one is talking about
AI agents capable of taking actions could also introduce a whole new set of security issues. What do they need to have access to? What should they be allowed to do? What if application-level permissions aren’t enough?
In this AI Stage session, “Agent Security Issues Nobody’s Talking About,” Okta President of Products and Technology Rick Smith and NanoCo Co-Founder and CEO Gabriel Cohen address agent security at the infrastructure level, including weaknesses in application-level permission models and architectural decisions that founders need to consider when deploying agent AI.
If agents are part of your product roadmap, security is also part of it. Get your tickets now to save up to $200 and add this session to your Disrupt agenda.
3. Securing the AI enterprise: Why the cloud has become more complex
Introducing AI products into your enterprise requires more than an impressive demo.
Security, governance, and observability are all part of the conversation when companies consider deploying AI to their critical systems.
In this AI session, “Securing the AI Enterprise: Why the Cloud Has Become More Complex,” Rudy Mitra, Vice President of AWS Security Services, Katie Moussouris, CEO of Luta Security, and cybersecurity veteran Wendy Nather explore the infrastructure that enterprises need as AI takes on a more autonomous role.
For founders building enterprise AI, this is your chance to hear what happens when your innovation meets the security requirements of the organizations you want as customers.
Do you want to build AI for your enterprise customers? Explore the security requirements that stand between your product and its deployment. Secure your pass to Disrupt by September 25th and save up to $200.
4. Building AI systems when failure is not an option
AI operating in the physical world increases risk. Mistakes don’t stop on the screen. It can impact vehicles, aircraft, industrial systems, and critical missions.
On the Real World AI Stage, Shield AI Chief Technology Officer Nathan Michael, General Motors Robotics Strategy Director Mikel Taylor, and Waabi Founder and CEO Raquel Urtasun bring defense and autonomous systems perspectives to one of the most difficult questions hard tech founders face: how do you determine if an autonomous system is safe to deploy?
The “Building AI Systems When Failure Isn’t an Acceptable” conversation will cover building a culture of safety, testing and validating AI, avoiding regulatory hurdles, and building companies that can earn trust when the consequences of failure are physical.
If failure is not an option, testing cannot be postponed. See how leaders in the self-government sector are working safely with Disrupt. Explore ticket options and save up to $200 on passes through September 25th.
5. The robot is waiting for the ChatGPT moment. Here’s what gets in the way
Robots have a data problem. They don’t have access to the massive training data pools that have helped accelerate progress in language models and self-driving cars.
This gap is one of the biggest obstacles to scaling physical AI. In this Real World AI session, “Robots are waiting for their ChatGPT moment, here’s what stands in their way,” Les Karpas, Global Head of Physical AI at NVIDIA Inception, explores how data pipelines, simulation environments, and underlying models can help solve this problem, and what it takes to build, test, and deploy more capable and reliable robots.
For founders working on physical AI, this session tackles the big questions. What will it take for robotics to reach its own ChatGPT moment and gain the trust needed to make these systems work in the real world?
Let’s take an inside look at what will help robotics clear the next big hurdle. Register with Disrupt today and save up to $200 before prices go up.
5 sessions. One question every AI founder needs to answer.
Can people trust what you’re building? These five conversations are part of more than 200 sessions across six industry stages, roundtables, and breakout sessions at Disrupt, October 13-15 at Moscone West in San Francisco. More than 10,000 founders, investors, executives, and technology leaders are expected to attend, more than 250 speakers, more than 300 startup exhibitors, and countless networking meetings.
You can build smarter models, more capable agents, or robots that do things no one has ever done before. But getting customers to trust it enough to adopt it may be one of the most difficult parts of bringing it to market.
Secure your pass to Disrupt now and save up to $200 by September 25th at 11:59pm PT. Your second ticket will be 50% off. Hear how AI leaders are tackling the safety, security, and reliability challenges that stand between breakthrough technology and real-world deployment.
If you make a purchase through links in our articles, we may earn a small commission. This does not affect editorial independence.
