At its annual Relate conference, Zendesk has unveiled a strategic shift in its AI roadmap, introducing what it calls the Autonomous Service Workforce.
The move signals an evolution away from traditional, deflection-based chatbots toward specialized AI agents designed to handle complex service tasks across various channels and use cases.
The new framework is built upon the Zendesk Resolution Platform, a unified system that integrates data, intelligence, and workflows. To power these capabilities, Zendesk is leveraging a Resolution Learning Loop, which utilizes insights from approximately 20 billion ticket interactions to refine automated responses and bridge knowledge gaps in real time.
Central to this announcement is a shift in how enterprises will deploy and pay for AI. Zendesk is introducing a new Agent Builder, omnichannel AI agents, and Copilots for human staff. Perhaps most significantly, the company is moving toward outcome-based pricing, where enterprises are charged based on verifiably resolved issues rather than simple interactions or seat licenses.
Tom Eggemeier, CEO of Zendesk, emphasized that this transition marks a fundamental change in the relationship between AI and human agents.
“The era of the chatbot – the era of frustration and deflection – is over. We are entering the age of the Autonomous Service Workforce. We believe every business will soon run on specialized AI agents that work alongside human experts as one unified team. These agents will be more than just code; they will be team members, held to the same high standards of accountability as any human. Our vision is to put the power to build this workforce into the hands of every enterprise, on one elegant platform. Whether those agents are crafted by Zendesk, by our partners, or by your own teams, they will all speak with one voice. We are providing a future where AI is the foundation, and human experts are the architects.”
The strategy aims to rectify a long-standing issue within the CX industry where disconnected tools and legacy workflows often prioritized ticket deflection over actual problem-solving. By treating AI agents as accountable team members, Zendesk seeks to provide a more cohesive service experience.
This transition is powered by new capabilities that turn service into a coordinated, self-improving operation. Key updates for the agentic era include:
Agent Builder and Custom Agents
- Zendesk introduced Agent Builder, a no-code interface that enables companies to build, test, deploy, and optimize custom AI agents tailored to their unique policies, workflows, data, and business logic. Agent Builder is designed to help organizations automate more complex front, middle, and back office service work while maintaining governance and oversight from a single control plane.
Expanded Zendesk AI Agents
- Zendesk AI Agents now operate across messaging, email, voice, and AI platforms like ChatGPT and Gemini, maintaining shared context and continuity across interactions. Zendesk has expanded this functionality in part through its acquisition of Forethought. These agents are designed to operate across Zendesk and external service environments – including through the new Zendesk MCP Server, which connects Zendesk tickets, knowledge, and customer data to external AI platforms so brands can deliver personalized, verified answers wherever customers seek help.
- Zendesk also announced expanded support for Voice AI Agents, including multi-brand and multilingual support across more than 60 languages. The agents can now switch languages mid-conversation while maintaining full context and continuity, strengthening Zendesk’s CCaaS offering powered by Amazon Connect.
AI Agents for Employee Service
- Zendesk announced fully autonomous AI agents for Employee Service, purpose-built for internal support. Powered by Zendesk’s Unleash acquisition, these agents operate in tools such as Slack and Microsoft Teams, search across enterprise systems, and enforce source-level permissions so employees only receive answers they are authorized to access.
New Copilot Experiences for Agents, Admins, Knowledge Teams, and Analysts
Zendesk announced major advancements across its Copilot portfolio to support human teams:
- Agent Copilot: Designed to work out of the box by connecting to internal and external sources to generate procedures and take action on at least 30% of tickets from day one.
- Admin Copilot: Currently available, this tool helps administrators identify operational issues, recommend fixes, and apply workflow changes in real time.
- Knowledge Copilot: Now in early access, this copilot identifies gaps, outdated content, and inconsistencies based on real customer conversations.
- Analyst Copilot: Also in early access, this assistant helps teams spot trends and understand root causes through a new agentic analytics experience.
Quality Score for Continuous Quality Assurance
- Zendesk announced Quality Score, which brings automated, continuous quality measurement to Suite Professional plans and above. Quality Score analyzes 100% of human and AI interactions, giving teams an objective view of service quality and surfacing opportunities to improve performance in real time.
Context Graph and Expanded Knowledge Graph
- Zendesk introduced Context Graph, the operational memory layer that powers the company’s agentic analytics experience by capturing past analyses, agentic reasoning, and performance context to improve future recommendations. Zendesk also announced expanded Knowledge Graph connectors, supporting sources such as SharePoint, Google Drive, Notion, Guru, Contentful, Document360, and more.
Action Flows for AI Agents and New Workflow Connectors
- Zendesk announced Action Flows for AI Agents, allowing teams to build integrated workflows for AI agents directly inside Action Builder. This capability helps AI agents take action across systems with greater governance and orchestration. Zendesk also announced a new library of 40 prebuilt workflow connectors, for systems such as Okta, Claude, and OneDrive, with more than 100 additional apps planned by the end of the year.
Zendesk Model Context Protocol
- Zendesk announced support for Model Context Protocol (MCP), including client and server experiences. With Zendesk MCP Client, AI Agents and Agent Copilot can connect to external systems once and automatically expand their capabilities as new MCP tools are added. With Zendesk MCP Server, businesses will be able to connect Zendesk tickets, knowledge, and other data to external AI systems in a trusted and governed way.
A New Business Model for the Workforce Era
- Zendesk is also expanding its outcome-based pricing model this year. Every resolution Zendesk charges for is verified — both by the AI agent resolving the interaction end-to-end and independently confirmed by a dedicated AI evaluation model. Spam and routine exchanges are excluded. Customers pay only for outcomes that are genuinely, confirmably resolved.
Daniel Newman, CEO, Futurum Research, commented:
“What’s compelling about Zendesk’s direction is that it recognizes a core truth about service: automation on its own is not enough. To improve the experience meaningfully, AI has to be part of a broader system that can connect context, take action, and evolve with the needs of the business. That’s the kind of approach that can help organizations build a more scalable and responsive support experience over time.”
Zendesk’s pivot toward an Autonomous Service Workforce represents a maturation of the AI narrative within the CX industry. By moving away from deflection—a metric often criticized for prioritizing cost-cutting over customer satisfaction—and toward resolution, Zendesk aims to align its technology with the actual goals of the end-user.