Adobe Deploys AI Agents to Automate Marketing Workflows

Adobe announced an expanded, production-ready agentic AI stack that brings autonomous, multi-agent capabilities to marketing and customer experience orchestration. The announcement highlights the Agent Orchestrator in Adobe Experience Platform, new enterprise packaging under Adobe CX Enterprise, an AI coworker surfaced as AI Assistant, plus expanded integrations so agents and developer tools run across major cloud vendors. Adobe reports over 70% of eligible AEP customers are using the conversational assistant as an entry point for agents.

Agent Orchestrator is a reasoning and orchestration layer that selects, composes, and sequences purpose-built agents using enterprise context from Adobe Experience Platform. Key technical elements include MCP endpoints for model-context plumbing, agent skills that encapsulate discrete capabilities, and a reasoning engine that can build step-by-step plans and adapt based on feedback. Agents operate with product-level access controls and consume AI credits when performing agent jobs, enforcing usage and governance limits. Practitioners should note these operational features:

• •Agents remember conversation history and can combine outputs from multiple agents to create unified responses.
• •The orchestrator exposes a reasoning panel, enabling visibility into the stepwise plan and which agents executed which tasks.
• •Integration points include SDKs and connectors that let agents act across Real-Time CDP, Journey Optimizer, Experience Manager assets, and analytics surfaces.

This is Adobe moving agentic AI from pilots into enterprise-grade workflows. Unlike single-call copilots, the stack targets multi-step orchestration across marketing stacks where decisions depend on customer data, content, and channel rules. By tying agents directly to AEP, Adobe leverages existing customer graphs, identity resolution, and content stores to keep actions contextually relevant and auditable. That data-first approach differentiates Adobe from vendors offering standalone LLM copilots that lack deep enterprise data integrations. The partner expansions also signal a pragmatic strategy: let brands run agent-enabled experiences on AWS, Azure, Google Cloud, and other surfaces while keeping orchestration and governance in AEP.

Adoption will hinge on three practical questions for teams: how AI credits map to real costs for common agent jobs, how access controls and audit logs meet compliance needs, and how well the reasoning engine avoids cascading automation errors in production. Expect more customer case studies from B2B and B2C implementations, and incremental feature releases that broaden agent skills and partner-hosted execution templates.

"Our agentic AI innovations are elevating customer experience orchestration by reimagining processes, unlocking productivity for marketing teams and delivering personalized experiences at scale to drive growth," said Anjul Bhambhri, senior vice president of engineering, Adobe Experience Cloud. Cisco and other early enterprise users report practical gains in shortening buying cycles and improving account engagement by combining engagement insights with agentic automation.

Overall, this is a move from experimentation to operationalized agentic AI for marketing and CX. Teams evaluating platform choices should benchmark not only generative quality, but orchestration reliability, integration depth with customer data, and the operational model for cost and governance.