Agentic AI has moved from theoretical promise to operational reality across experience management platforms, with organisations now deploying autonomous agents to handle everything from routine inquiries to complex multi-step customer journeys. The shift reflects a fundamental change in how CX teams architect their operations: rather than using AI as a support layer within existing workflows, teams are increasingly designing workflows around what agents can autonomously execute. This represents a departure from the chatbot era, where human handoff remained the default endpoint. The critical question for teams already running Agentforce or similar platforms is whether their current governance and escalation frameworks can accommodate agents that resolve issues without human intervention—a capability that demands different oversight models than traditional automation.
The business case has solidified. Research shows 96% of organisations report that agentic AI deployments met or exceeded ROI expectations in 2026, signalling that the technology has moved beyond early adopter enthusiasm into mainstream viability. However, this adoption wave masks a structural tension: as Klarna's shift to gig workforce for customer service amid AI integration demonstrates, agentic AI doesn't simply replace headcount—it reshapes workforce composition and skill requirements. For support team leads, this means the conversation is no longer about whether to deploy agents, but how to redeploy existing talent toward higher-judgment work that agents cannot handle, and whether your organisation's hiring and training pipelines can adapt at the required pace.
The governance gap remains the most pressing operational challenge. Questions around who governs AI agents in contact centres highlight that technical deployment has outpaced policy frameworks. CX consultants and administrators must now establish clear accountability structures for agent decisions, particularly in regulated industries where autonomous resolution carries compliance risk. The platforms themselves—Adobe, Salesforce, and others—are building governance tooling, but implementation remains fragmented. Teams that establish governance early, before agent deployments scale, will avoid the costly retrofitting that early adopters are now navigating.
Real-world agentic AI use cases in experience management. Adobe for Business
Real-world agentic AI use cases in experience management. Adobe for Business