The second day of TechEx North America shifted toward a more in-depth, critical look at how AI is being used in businesses—though the overall tone remained hopeful. The AI and Big Data program kicked off by addressing what’s known as the “AI graveyard”: projects that show promise during testing but fail to deliver in real-world settings. Despite this somewhat gloomy label, several speakers and sessions focused on practical strategies that forward-looking companies can use to avoid ending up in that technological dead zone.
Across various tracks on day two, the event dug deeper into the widespread challenges affecting AI rollouts. Sessions in the Enterprise AI Implementation, ROI, and Adoption tracks used stalled pilot projects as a springboard to explore why these initiatives often fall short. Attendees received valuable guidance, including advice on targeting agentic AI toward specific business functions, preparing data infrastructure to support agents effectively (laying the groundwork for success behind the scenes), and understanding how token-based pricing models impact a company’s bottom line.
On the infrastructure side, there were also detailed discussions about whether organizations should purchase or build their own physical hardware for AI initiatives—and how to generate lasting, meaningful returns from data and AI investments when all influencing factors are carefully weighed.
In cases where AI deployments stall, the root problem often comes down to the idea of the “personal copilot.” This works well for individual employees and their personal workflows, but struggles to scale across an entire department—let alone a whole organization. Many firms have the budget to launch small-scale AI experiments at the user level, and these often yield impressive results. When the user is a C-suite executive, the efficiency gains tend to spark excitement company-wide, which is certainly a positive sign. However, moving from isolated wins to organization-wide transformation is where most businesses hit roadblocks. This challenge formed the core focus of day two’s discussions across the show floor and multiple stages at the San Jose McEnery Convention Center.
Cybersecurity challenges
Even with terms like “stalled” and “hard to scale” in play, speakers on the Cyber Security and Cloud Expo stage highlighted the rapid pace at which companies are adopting agentic AI as a key driver of a “velocity gap.” When AI deployments succeed, they gain momentum quickly—but security and governance issues arise when business units adopt generative AI faster than security teams can manage or safeguard it.
Like a classic double-edged sword, AI is reshaping both offense and defense in cybersecurity. Internally, risks emerge from uncontrolled agents and large language models, while externally, attackers now wield AI-powered scanning tools that can detect vulnerabilities more efficiently.
Another recurring theme in keynotes and roundtables was the modern evolution of shadow IT—now called shadow AI. If employees input sensitive data into unapproved tools, or if sanctioned AI systems lack proper boundaries and oversight, the organization’s attack surface can expand without the cybersecurity team even noticing. As a result, data governance and system monitoring are becoming more tightly linked than ever—a message echoed across both cybersecurity and Cloud/Big Data sessions.
For dedicated cybersecurity teams, zero trust was presented as a solution to uncontrolled AI adoption outside their oversight. This approach enforces a “deny by default” policy for both humans and machines. Identity verification and access controls must also apply to services and AI agents, ensuring automated workflows follow the same permission rules as every other component in the IT environment.
Day two of TechEx North America wasn’t about dampening leaders’ AI ambitions—speakers, thought leaders, and attendees alike treated AI and agentic systems as established realities. However, representatives from diverse industries and functions shared nuanced perspectives, blending concerns with optimism. Each contributed insights that enriched the broader conversation around AI implementation in 2026.
The rise of robots
Excitement was still palpable across many areas of the conference. Humanoid robots drew enthusiastic crowds (who doesn’t love a friendly android?), but more practically, the new Physical AI track attracted some of the largest audiences. Many attendees outside the track noted that software coding has been the first professional domain to see clear benefits from large language models. There’s also a growing consensus that automated physical systems will be the next major sector to gain from focused development of new AI models and their real-world applications.
The AI models powering next-generation physical AI are unlikely to be LLMs—though LLMs may still play a role if devices need to interact with humans. As these models mature and move beyond research, the TechEx Events series will be among the first to showcase them and demonstrate how they can operate effectively in business environments.
New learning opportunities at the event
This year’s event introduced a strong emphasis on hands-on coding, with interactive sessions guiding attendees through building their own AI agentic models—including techniques for enabling agents to self-improve—all using live Google Colab environments. The TechEx Learning Hub also hosted workshops led by Nvidia and the ever-popular Google Hackathon, catering to participants ranging from IDE beginners to experienced developers. Turning knowledge into action is central to the event’s mission, whether it’s C-suite leaders learning strategic best practices or developers bringing innovative ideas to life.
TechEx takes cutting-edge technology and filters it through a business-focused lens—practical yet forward-looking. Don’t miss the next edition of TechEx in Amsterdam this September. Who knows how much progress we’ll have made in just four months?
(Image source: TechEx Events)
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and co-located with other leading technology events. Click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
