Cisco Unveils Webex AI Agent Strategy to Bridge Collaboration and Customer Support

During the WebexOne event held in October 2026, Cisco introduced a series of advancements aimed at embedding intelligent agents into its communication ecosystem. Rather than treating artificial intelligence as a simple add-on, the company is focusing on agentic capabilities that allow these digital entities to function alongside human teams. The goal is to facilitate more fluid interactions across video conferencing, messaging, and customer support channels by providing agents with the necessary context to complete multi-part assignments.

A central component of this strategy is the introduction of a layer named Dialog, which is engineered to link internal enterprise databases with the Webex AI Agent framework. This system is designed to preserve the history and context of client interactions as they move through different stages, ensuring that follow-up tasks are handled smoothly once a conversation concludes. Furthermore, Cisco highlighted potential internal use cases where staff might summon specific agents within Webex spaces—such as a data agent to pull performance metrics or a creative agent to modify slide decks—streamlining tasks that typically require switching between multiple software applications.

Integration with external AI models is also a priority. The company revealed that agents powered by Anthropic's Claude are expected to participate directly in calls and meetings. Additionally, through the Model Context Protocol, Cisco plans to integrate Webex data into OpenAI's ecosystem. To manage the security risks associated with giving AI agents access to broad data sets, Cisco is launching AI Agent 360. This suite will utilize Cisco AI Defense and Splunk to provide monitoring and oversight, with the Agent Observability feature from Splunk scheduled for broad release in late 2026.

Technology administrators are encouraged to prepare for these changes by evaluating which specific business processes would benefit most from automation. While these tools offer the potential to minimize redundant data collection, Cisco emphasizes the importance of maintaining human oversight and strict access permissions. Initial testing through controlled pilot programs is recommended to verify how agents interact with sensitive company systems. While many of these features are currently in development, Cisco expects a beta rollout to begin in early 2027.

---

Read original at TechRepublic AI.