Agentic AI is beginning to move from experimentation into measurable enterprise operations, and Cisco is providing a notable example of what that transition can look like.
Cisco has reported that its AI systems independently resolved 145,000 customer support cases during fiscal 2026 without human intervention. The development demonstrates how AI agents are moving beyond assisting support employees and beginning to handle defined customer service workflows independently.
For enterprises exploring Agentic AI, the announcement highlights a broader shift: the value of AI is increasingly being measured not only by its ability to generate content or answer questions, but by its ability to complete business processes and deliver measurable operational outcomes.
From AI Assistance to Autonomous Support
Traditional customer service AI has generally focused on assisting human agents.
AI-powered tools can summarize cases, recommend responses, search knowledge bases, and provide information to employees.
Agentic AI takes this capability further.
AI agents can be given specific objectives and the ability to execute multiple steps required to resolve a customer issue. Instead of simply recommending an action, an agent can potentially perform the necessary workflow itself within defined boundaries.
Cisco’s reported 145,000 fully AI-resolved cases provide an example of this model operating at enterprise scale.
Cisco Is Using Agentic AI Across Customer Experience
According to Cisco CEO and Chair Chuck Robbins, the company is deploying generative AI and agentic systems throughout its customer experience organization.
The goal extends beyond support automation.
Cisco says AI is helping accelerate:
- Quote turnaround
- Support case resolution
- Customer service operations
- Customer renewals
This positions Agentic AI as more than a cost-reduction technology. It can become part of the broader customer lifecycle, connecting service efficiency with commercial outcomes.
145,000 Cases Without Human Intervention
The headline figure is significant because Cisco says these 145,000 support cases were resolved entirely by AI with zero human intervention during FY2026.
This illustrates an important evolution in enterprise AI adoption.
Organizations have traditionally used AI as a copilot, with employees reviewing recommendations before taking action.
The emerging agentic model allows AI to independently handle selected workflows when the task falls within predefined parameters.
This could enable customer service organizations to handle higher volumes without requiring a human employee to manually oversee every interaction.
However, the scale of automation also raises an important question: Are organizations measuring the right outcomes?
Efficiency Is Only Part of the Story
The number of cases resolved autonomously demonstrates operational efficiency, but enterprise Agentic AI success should not be measured purely by automation volume.
Businesses also need to understand whether AI is improving:
- Customer satisfaction
- First-contact resolution
- Customer retention
- Resolution quality
- Customer effort
- Long-term loyalty
CX Today notes that Cisco’s reported figures demonstrate the ability to automate support at scale, while providing less evidence about whether customers consistently experienced better outcomes.
This distinction will become increasingly important as enterprises move from AI pilots toward production-scale deployments.
Cisco’s Circuit AI Assistant
Cisco has also embedded its proprietary Circuit AI assistant across its operations.
Circuit runs on Cisco’s secure AI infrastructure, which the company says is designed to improve GPU utilization and automatically route tasks to the appropriate large language model.
This architecture highlights another important element of enterprise Agentic AI: the infrastructure supporting AI agents can be just as important as the agents themselves.
Organizations operating AI at scale need to manage model selection, computing resources, token consumption, security, and workload placement efficiently.
The Infrastructure Challenge Behind Agentic AI
The growth of AI agents is also influencing enterprise infrastructure purchasing.
Cisco reported that enterprise Nexus switch orders associated with AI deployments increased by more than 85%, while data-center networking orders grew by more than 35% year over year.
The figures demonstrate how AI adoption is creating demand beyond software.
As organizations deploy increasingly complex AI workloads, they require infrastructure capable of supporting the necessary compute, networking, security, and data movement.
Model Selection Is Becoming a Business Decision
As organizations deploy multiple AI agents, they may not rely on a single model for every task.
Different workloads can require different combinations of:
- Cost
- Performance
- Latency
- Security
- Accuracy
- Data residency
- Model capability
Cisco highlighted the growing importance of selecting the right model and deployment environment for individual use cases.
For enterprises, this means AI infrastructure strategy is becoming increasingly connected to business strategy.
Security Must Scale With AI
Greater AI autonomy also increases the importance of security.
Cisco reported that more than 1,500 customers purchased its new security products during Q4, taking the number of net-new customers since launch above 6,400.
The development reflects a broader industry trend: as enterprises deploy AI across distributed environments, they also need stronger security controls to protect the data, systems, identities, and infrastructure used by AI applications.
For Agentic AI, this becomes particularly important because agents may interact with multiple enterprise applications and access business information while completing workflows.
Enterprise Budgets Are Being Reprioritized
Despite strong interest in AI, enterprises are not necessarily expanding their technology budgets at the same pace.
Cisco CEO Chuck Robbins noted that customers are increasingly reprioritizing existing budgets as they invest in AI infrastructure and related technologies.
This creates a challenge for organizations.
Companies must balance spending on:
- AI infrastructure
- Cybersecurity
- Cloud
- Data platforms
- Networking
- Existing IT systems
- Customer experience technologies
The result could be a significant shift in technology investment priorities as organizations determine which AI initiatives deliver measurable business value.
What Cisco’s Experience Means for Agentic AI
Cisco’s experience provides several lessons for enterprises considering Agentic AI.
1. Start With Defined Workflows
AI agents are more suitable for processes where objectives, permissions, and expected outcomes can be clearly defined.
2. Measure Business Outcomes
The number of automated tasks matters, but organizations should also track customer satisfaction, retention, quality, and revenue impact.
3. Build Strong Infrastructure
Agentic AI requires scalable compute, networking, data access, and model-management infrastructure.
4. Make Security Part of the Architecture
AI agents should operate within clearly defined security and access boundaries.
5. Optimize Model Usage
Different tasks may require different AI models, making intelligent model routing and workload management increasingly important.
The Future of AI-Powered Customer Service
Cisco’s 145,000 autonomous support cases demonstrate that Agentic AI is moving toward a new stage of enterprise adoption.
The next phase will not simply be about deploying more AI assistants.
Instead, organizations will increasingly build AI-powered operational systems capable of completing defined processes from beginning to end.
Customer support is likely to be one of the most important areas for this transition because many service workflows are repetitive, data-intensive, and governed by clearly defined processes.
At the same time, human employees will continue to play an important role in complex cases, exceptions, escalations, and situations requiring judgment or empathy.
Conclusion
Cisco’s reported 145,000 AI-resolved support cases represent an important milestone in the evolution of enterprise Agentic AI. The company is demonstrating how AI can move beyond assisting employees to independently completing defined customer service workflows.
But the larger lesson is that successful Agentic AI adoption requires more than autonomous agents.
Enterprises need the right combination of data, infrastructure, security, governance, model management, and measurable business outcomes.
As AI agents become increasingly capable of executing real-world workflows, the competitive advantage will shift toward organizations that can deploy them safely, efficiently, and at scale.
The future of enterprise customer experience may therefore be less about AI helping employees do their jobs—and more about AI agents becoming active participants in how the business operates.
