Every enterprise is asking the same question: How do we move from AI experimentation to AI at scale? The technology is evolving rapidly. Employee expectations continue to rise. Business leaders want measurable outcomes, not isolated pilots.
Yet many organizations are still trying to layer AI onto systems and processes that were never designed for it. At Cisco, we’ve found that operationalizing AI isn’t simply about deploying more models. It’s about making AI part of how the business works every day.
We’ve learned a lot over the past two years. While the 2025 Cisco AI Readiness Index found that only 33% of organizations have a formal plan to guide employees through AI adoption, we took a different path. By focusing on three ideas, in particular, trusted data, one secure platform, and AI-native workflows, we continue to shape how we approach that challenge.
Connecting AI to Trusted Enterprise Data #
AI is only as good as the data it can access. That sounds obvious, but it’s one of the biggest challenges enterprises face. Business data is spread across applications, warehouses, documents, and legacy systems. Even the best model won’t deliver useful answers if it can’t securely access the right information or understand the context behind it.
That’s why we started with our data. We’ve invested in bringing enterprise data together responsibly, connecting AI to the applications where information already lives, and building the semantic understanding needed for AI to reason across the business.
Trusted data isn’t just an input to AI. It’s what allows employees to trust the answers that AI gives them.
Defeating Shadow AI with a Secure Platform #
When generative AI first emerged, we witnessed the same behavior as many organizations. Employees immediately began experimenting with consumer AI tools. We recognized early that if we didn’t provide a secure alternative, shadow AI would become the default. We had two choices: try to stop it or give employees a better alternative. We chose the second option.
That decision led us to build Circuit, our secure, responsibly governed, multi-model agnostic platform. What began as a secure way for employees to access leading AI models has become the front door to AI at Cisco and the platform through which we’re operationalizing AI across the company.
Employees can access multiple AI models through a single trusted experience, connect AI to enterprise data, share prompts and projects, build connectors and agents, and increasingly automate work – all without switching between disconnected tools.
Three principles shaped how we built Circuit: it had to be secure, compelling, and extensible. Secure enough for employees to confidently work with enterprise data and aligned with our Responsible AI Principles. Compelling enough to give them access to the right model for the right task. And extensible enough for teams to build and share prompts, projects, connectors, agents, and reusable capabilities.
Those principles helped Circuit reach more than 100,000 users and 90% employee adoption. More importantly, it provided the foundation for a culture where employees built and shared capabilities of their own. We didn’t get there through mandates or usage targets. We made AI secure, useful, and easy to experiment with – and adoption followed.
Becoming AI Native: Redesigning the Workflow, not just the task #
Building a platform is only the beginning. The biggest opportunity is redesigning work around what’s now possible. Too often, organizations take a ten-step process and use AI to improve each step. We’ve found the better approach is to rethink the experience from the beginning. That’s what we mean by becoming AI-native.
Across Cisco, employees use AI every day to summarize information, analyze documents, search internal knowledge, generate content, and automate routine work. More than 21,000 Cisco engineers now use AI coding tools, with more than 80% using them every week. Engineers are saving an average of six hours every week, while employees across the broader business are saving an average of five. The time savings matter.
The bigger impact is that AI is reducing friction. People spend less time searching for information and moving between systems, and more time solving problems, making decisions, and creating value. That’s when AI stops being another technology initiative and starts becoming part of how the enterprise operates.
The next step is agentic AI. As AI moves beyond answering questions to completing work, the platform becomes even more important. Agents need trusted data, secure access, enterprise context, and clear governance. The goal isn’t maximum autonomy, it’s the right level of autonomy for the right task.
**Why operational excellence wins the AI race **
Operationalizing AI is about creating the conditions for AI to become a trusted part of how the enterprise operates.
We’re still learning. The technology will continue to evolve, and so will our approach. But one thing has become clear. The organizations that create the most value from AI won’t necessarily be the ones that adopt it first. They’ll be the ones that operationalize it best.
Visit the #
[Cisco AI Readiness Index]to benchmark your organization’s progress and access resources to help guide your own AI journey.