• New virtual agent capabilities reduce customer effort, prevent repeat contacts, and give service leaders confidence to scale automation
  • With 43% of consumers saying chatbots fail to resolve their issues, Zoom Virtual Agent 3.0 can help organizations close that gap

SAN JOSE, Calif., Feb. 24, 2026 (GLOBE NEWSWIRE) -- Today Zoom Communications, Inc. (NASDAQ: ZM) unveiled Zoom Virtual Agent 3.0 (ZVA), the next evolution in agentic automation. ZVA introduces a new execution architecture and expanded AI capabilities designed to resolve customer issues end to end, seamlessly hand off to human agents, and help enterprises shift from transactional service interactions to connected customer relationships.

Organizations face growing pressure to automate more customer service interactions as volumes rise and cost efficiency becomes a priority. Enterprises are entering what Zoom calls the “resolution economy,” where competitive advantage is defined not by speed, but by first-contact resolution, reduced repeat contacts, and end-to-end workflow completion. Yet disjointed virtual agents can get stuck in silos, unable to seamlessly transfer full context to human agents, creating bottlenecks for complex issues. A recent Morning Consult report commissioned by Zoom found that the top three chatbot frustrations among the groups surveyed are failure to resolve the issue (43%), getting stuck in a loop (38%), and having to repeat information (37%).

“Agentic AI was just the beginning,” said Chris Morrissey, general manager of Zoom CX. “Zoom Virtual Agent 3.0 orchestrates multi-step workflows across systems, continuously learns from human resolutions, and provides full transparency into every agentic action. This allows organizations to confidently automate complex interactions. It’s more than a product update, it’s another step toward more connected customer relationships, where AI and humans work together to resolve issues faster and build trust.”

What's new in Zoom Virtual Agent 3.0

The next evolution of Zoom Virtual Agent operates across voice and chat, and introduces a new execution architecture designed to automate complex, cross-system interactions with enterprise governance. Unlike competing solutions focused primarily on conversational containment, ZVA is built for execution and resolution, securely orchestrating multi-step actions across systems with observability, control, and measurable outcomes. The following new features are now available:

  • Enhanced AI execution framework: Zoom Virtual Agent 3.0 is built on the latest Zoom AI Companion 3.0 architecture, allowing multi-step workflows to run across compatible CRM, billing, order management, and other enterprise systems with full observability and governance. This extends Zoom’s agentic AI capabilities by enabling full-cycle resolution rather than just single-step responses.
  • Agent journey transparency and governance enhancements: Account admins can now see the data sources, decision logic, and workflow paths behind automated actions. CX teams can audit performance, troubleshoot breakdowns, and refine automation policies, enabling responsible scaling without sacrificing adherence or control.

The following next-generation enhancements, expected to be generally available in Spring 2026, will expand ZVA’s ability to handle more complex workflows while improving reliability, oversight, and operational efficiency:

  • Multimodal large language model (LLM) intelligence: ZVA will be able to interpret and act on customer-submitted documents, images, and structured identifiers such as serial numbers and forms. By extracting relevant data directly from visual and document inputs, the virtual agent can automate service scenarios that previously required manual review, helping reduce customer effort and the need for escalation.
  • Continuous learning: When integrated with Zoom Contact Center, ZVA extracts insights from escalated engagements that human agents successfully resolved and applies those validated recommendations, with oversight and controls, to similar future requests. This creates a structured feedback loop based on human-agent resolutions, reducing repeat contacts and improving resolution consistency over time.
  • Proactive outbound engagement: The virtual agent can initiate contact, confirm updates, and complete tasks based on known events. This helps organizations resolve issues before customers reach out, reducing inbound volume and customer effort.

Zoom Virtual Agent in action

Modern customer issues rarely fit into a single scripted response. Zoom Virtual Agent (ZVA) is designed to move beyond basic automation, including authenticating users, interpreting inputs, orchestrating backend systems, and completing real business actions within a single, connected workflow.

By reducing repetitive steps and minimizing handoffs, ZVA helps increase first-contact resolution while lowering handling time and operational effort. An example of this is below:

  • Scenario: End-to-End Warranty Fulfillment
    • When a customer submits a warranty claim, ZVA can authenticate the user, extract a serial number from an uploaded image, validate eligibility across backend systems, schedule device pickup, initiate a replacement order, and confirm shipment, all within one continuous interaction.
    • If escalation is required, the complete workflow history, including verified inputs and actions already taken, transfers seamlessly to a live agent. The customer does not need to repeat information or restart the process, accelerating resolution and improving the overall experience.

Raising the bar for resolution

As organizations automate more interactions to manage rising volumes and cost pressures, performance is increasingly measured by resolution quality — not containment rates alone. In the resolution economy, automation must accurately interpret intent, execute across systems, and meaningfully reduce operational load.

Zoom is already seeing measurable results internally after implementing the latest updates to ZVA in its own virtual agents:

  • Query understanding accuracy: Zoom’s no-match rate (the percentage of total conversation turns in which a virtual agent failed to understand the user's input) has dropped from 35% to 0%, meaning almost all customer requests are accurately interpreted on the first attempt, reducing repeat queries and friction.
  • Significant time savings: On Zoom’s billing team, deflection rates (measuring the percentage of customer support inquiries resolved through self-service tools) rose from 0% to 30% in just three months, saving over 1,000 agent hours per month.

Zoom Virtual Agent 3.0 reflects Zoom’s broader ambition to power intelligent, connected customer experiences where AI and human agents work together to complete complex interactions with speed, transparency, and trust. This is automation designed not just to scale, but also to earn trust with every interaction.

See Zoom Virtual Agent in action

Read the Zoom blog to see how ZVA supports Zoom’s own customers by improving resolution at scale, and stop by Zoom’s booth (#519) at Enterprise Connect 2026, March 10–12 in Las Vegas, to see a live demo.

About Zoom

Zoom (NASDAQ:ZM) provides the AI-first, open work platform built for human connection and purposefully designed to move conversations to completion. From entrepreneurs to global enterprises, customers choose Zoom to seamlessly collaborate, communicate, and drive outcomes across meetings, chat, phone, contact center, events, and more — all with the built-in assistance of Zoom AI Companion. Founded in 2011, Zoom is headquartered in San Jose, CA. For more information, visit zoom.com .

Zoom Public Relations
Travis Isaman
[email protected]