Cisco is expanding Webex with collaborative AI agents designed to work across applications, manage longer-running tasks, and support employees beyond traditional prompt-based assistants.
Announced at WebexOne 2026, the updates extend Cisco’s agentic AI strategy across collaboration, physical workplaces, and customer experience. They include new Webex AI agents, Intelligent Workplace Experiences, RoomOS 27, and Dialog, an agentic CX platform designed to maintain context and complete work across systems over time.
For channel partners, the broader shift could create new opportunities around AI deployment, application integration, governance, workplace modernization, and customer experience as enterprises decide how much autonomy to give agents operating across their environments.
Webex adds AI agents that work across applications
Cisco is moving Webex beyond prompt-based AI assistants with collaborative AI agents that can be added directly to spaces, meetings, and calls.
“What this means is really people and agents working together is how collaboration is going to be done going forward,” said Amit Barave, vice president and general manager of Webex Suite, at a recent Cisco press briefing.
The agents are designed to execute multi-step work across Webex and third-party applications. Cisco gave the example of a user “@mentioning” an analytics agent in a Webex space to retrieve usage statistics, then involving a presentation agent to update a PowerPoint deck with the results.
“We are bringing in managed agents to Webex, which will also be able to do work, long-running tasks, get work done overnight, and come back to you for your approval,” Barave said.
Cisco is also introducing a digital chief of staff that can surface items requiring attention while maintaining context around a user’s work.
“When you have agents working overnight and doing things they need to do, the user experience or the experience in the platform has to be fundamentally redefined and reimagined,” Barave said. “And that is what Webex is becoming. It’s evolving into your work AI.”
The company said identity, security, governance, and visibility controls will determine what AI agents can access, what they can do, and how they operate within enterprise environments.
RoomOS 27 expands Cisco’s intelligent workplace strategy
Cisco is also introducing Intelligent Workplace Experiences, bringing secure networking, collaboration devices, digital experience assurance, and physical workplace systems into a common architecture managed through Cisco Cloud Control.
“We are making every room a space that can sense what is happening, reason about the needs of the team and act to support them,” said Espen Loberg, vice president and general manager of Cisco Devices. “We’re moving from managing workspaces, to managing the entire floor plan, to engineering a fully intelligent workplace.”
As part of the rollout, Workplace Designer is expanding beyond individual meeting rooms. IT and facilities teams can use a street address or an uploaded floor plan to create designs that cover networking, collaboration, and workplace intelligence ahead of deployment.
Cisco also announced RoomOS 27, the next version of its operating system for collaboration devices.
New capabilities include:
- Native Google Meet support: Cisco devices will support fully featured Google Meet alongside Webex, Microsoft Teams, and Zoom.
- Cisco AI Assistant: Users will be able to interact with meeting rooms using natural language and ask follow-up questions in 32 languages.
- Director Agent: The agent coordinates cameras, microphones, and audio during meetings.
- Expanded IT visibility: Workspace Advisor, Smart Room Assurance, Divisible Rooms, Splunk, and agent integrations provide additional tools to monitor and manage workspaces.
RoomOS 27 is expected to become available in November.
“As AI moves into our physical rooms, security cannot be bolted on after the fact. It must be built into the foundation,” Loberg said.
Cisco also introduced Cisco Ceiling Speaker Pro, an IP-connected ceiling speaker that integrates with Cisco cameras, microphones, and room devices and uses AI to adjust audio across meeting spaces.
Dialog brings longer-running AI agents to customer experience
For customer experience teams, Cisco is introducing Dialog, an agentic harness built for the Webex AI Agent platform.
Dialog is designed to maintain customer context across interactions, channels, systems, and over time, rather than treating each engagement as a separate conversation.
One of its main capabilities, long-horizon execution, allows agents to continue working after a customer interaction ends, coordinating across people, other agents, and backend systems until a requested outcome is completed.
Dialog also includes:
- Persistent relationship context: Combines customer history with enterprise data across channels and interactions.
- Self-learning with every interaction: Refines agent behavior by leveraging customer journeys and human-agent judgment while operating within enterprise policies.
- Rapid onboarding: Lets organizations onboard agents using existing conversations, knowledge bases, and operating procedures.
Cisco said Webex AI Agents will operate with active safety guardrails from Cisco AI Defense, while live telemetry and operational observability will connect with Splunk.
Webex expansion creates new integration work for partners
Cisco’s latest Webex updates extend agentic AI across collaboration, workplace infrastructure, and customer experience, increasing the number of systems partners may be asked to connect and manage.
“Just imagine if you always had an AI director in the meeting room with you. You would always get that best view,” said Snorre Kjesbu, senior vice president and general manager of Cisco Collaboration. “When we have collaboration extending beyond human participation, that means that we actually use those agents and those solutions as well.”
For MSPs, that shift could create additional work around AI deployment, governance, workplace modernization, and CX integration, particularly for customers already using Cisco networking, collaboration, security, or observability products.
It also raises operational questions about permissions and the level of autonomy organizations want to give AI agents as they take on more work across applications.
“We can share that context across people, across applications, collaboration agents, and then still keep human judgment and accountability as key to all of this,” Kjesbu said.
That could give partners a larger role not just in deploying these capabilities, but in helping customers determine where agents can act independently and where human approval or additional controls are required.
Meta added new small business workflows to its Muse AI agent, expanding integrations across tools such as Canva, Notion, and QuickBooks. Read more about what MSPs should assess around app integrations, permissions, costs, and oversight.