The next phase of A.I. adoption hinges less on compute and more on whether sensitive data can be protected in use. Unsplash

Generative A.I. has long been treated like a public experiment. Every week, a new model is designed. Yet according to industry experts, A.I. is advancing faster than the trust required for it to scale. The sector has succeeded in training machines to perform increasingly complex tasks, but the data that powers these systems remains too sensitive to surrender. As a result, the central question is no longer whether A.I. can perform, but whether it can be trusted to handle sensitive information responsibly, which has been one of the primary reasons enterprises remain cautious about full adoption.  

Sign Up For Our Daily Newsletter

Thank you for signing up!

By clicking submit, you agree to our terms of service and acknowledge we may use your information to send you emails, product samples, and promotions on this website and other properties. You can opt out anytime.

This is why Confidential A.I., which uses confidential computing—a security technology that protects sensitive data while it is being processed in memory—is not an experimental innovation. The success of A.I. adoption strongly depends on it. As the shift takes hold, 2026 will mark the year A.I. breaks from theory to infrastructure, from optional to essential. 

Trust as a major barrier to A.I. Adoption

Enterprise deployment is already revealing a key pattern. McKinsey’s 2025 global A.I. survey shows that 88 percent of organizations are now using A.I. in at least one business function, up from 78 percent just a year earlier. By that measure, the A.I. revolution is already well underway.

But a closer look tells a more complicated story. The same data shows that only one-third of these organizations have successfully integrated and scaled A.I. across the enterprise. Most remain stuck in pilot mode, held back by........

© Observer