Cisco Identifies Three Priorities for Scaling AI Across the Enterprise
As organisations move from AI experimentation towards enterprise-wide deployment, Cisco is highlighting three priorities it says will determine whether businesses can turn AI investments into measurable operational value: trusted enterprise data, secure AI platforms and AI-native workflows.
The shift comes as businesses increasingly look beyond individual AI tools and pilots and seek to integrate AI into everyday processes. However, many organisations are attempting to add AI to legacy systems and workflows that were not designed for increasingly autonomous technologies.
According to Cisco’s 2025 AI Readiness Index, only 33% of organisations have a formal plan to guide employees through AI adoption, highlighting a gap between AI experimentation and structured enterprise deployment.
Building AI on trusted data
Cisco argues that reliable enterprise data is the foundation for useful AI. Business information remains distributed across applications, data warehouses, documents and legacy systems, making secure access to relevant information and context critical to producing reliable AI outputs.
The company recommends connecting AI to existing enterprise applications and creating the semantic context needed for systems to reason across different sources of business information.
Providing a secure alternative to shadow AI
The rapid adoption of consumer generative AI tools has also created challenges around shadow AI, as employees experiment with services outside formal IT and security controls.
Cisco said its approach has been to provide employees with a secure, governed alternative through an internal, model-agnostic AI platform. The platform is designed to combine security and responsible AI controls with access to appropriate models and the ability for employees to develop and share prompts, projects, connectors and AI agents.
Rethinking workflows rather than individual tasks
Cisco’s third principle is to redesign entire workflows rather than simply applying AI to individual tasks within existing processes.
The company said more than 21,000 of its engineers use AI coding tools, with engineers saving an average of six hours per week. Across the broader business, employees are saving an average of five hours per week.
Cisco argues that the greater opportunity extends beyond productivity gains, allowing employees to spend less time searching for information or navigating between systems and more time making decisions, solving problems and creating value.
As AI systems move from answering questions towards completing tasks, the company said organisations will need stronger foundations around data, security, enterprise context and governance. The focus, it argues, should not be maximum autonomy, but the appropriate level of autonomy for each task.
For businesses looking to scale AI, the challenge is therefore shifting from simply adopting the technology to building the infrastructure, governance and workflows required to operate it effectively across the enterprise.
