Cisco Puts Claude Agents Inside Webex: Your Next Meeting Attendee Is an AI
Cisco unveiled Claude Managed Agents inside Webex at WebexOne 2026 — AI agents that join meetings live and execute tasks instead of just summarizing them.
The announcement
At WebexOne 2026 in Austin on October 7, Cisco unveiled Claude Managed Agents inside Webex — AI agents built on Anthropic's technology that join meetings, spaces, and calls as working participants. You can @mention an agent mid-meeting: ask an analytics agent to pull usage numbers, then have a presentation agent drop those figures into a PowerPoint deck, live, without anyone opening the file.
The agents handle multi-step tasks across Webex and outside applications, drawing on shared enterprise context already sitting inside the platform — scheduling, task tracking, and content generation from within the conversation.
From summary bots to participants
This is a meaningful shift. Workplace AI has spent two years as reactive: summarize the transcript, draft the follow-up email, log the action items — always after the fact. Cisco is describing agents that act during the work: standing participants in the team's workflow that you can hand a task to mid-conversation, the way you'd delegate to a colleague.
The infrastructure behind it matters too. Claude Managed Agents are cloud-hosted agents designed for background, multi-step work inside group conversations — closer to an active team member than a bolted-on feature.
The competitive picture
Every collaboration platform is racing to the same destination. Microsoft is grounding Copilot agents in business data, Google is wiring Gemini into Workspace, and now Cisco is putting agents inside the meeting itself. The winner will be decided by which agents can actually execute across tools — not which demos best.
What this means for your business
- Meetings become work, not just talk. If agents can pull data and build deliverables during the call, meeting notes stop being the output — the finished work is. Rethink your meeting workflow around live execution. - Delegate to agents like colleagues. The @mention model lowers the adoption barrier: no new app, no training course. Start by assigning one agent a recurring meeting task — weekly numbers, status decks — and measure the hours saved. - Governance follows the agents. Agents with access to enterprise context and outside apps need the same controls as employees: permissions, audit trails, and clear boundaries on what they can touch.