Adobe Brings Its Marketing Agent to ChatGPT, Claude and Copilot
IA en un minuto newsroom · Editor: Jon Elgezabal
In 30 seconds
Adobe lets teams work with CX Enterprise Coworker, its agent for campaigns and buying journeys, from Databricks Genie, Gemini for business, Claude, ChatGPT and Microsoft Copilot. Snowflake, IBM and Amazon Quick will follow soon, and there are new MCP connectors for Meta, Demandbase or Amazon Ads. Adobe Experience Platform is essential, together with Customer Journey Analytics, Journey Optimizer or Real-Time CDP.
On Wednesday, November 4, Adobe said that CX Enterprise Coworker, its AI agent for customer experience (marketing, campaigns and buying journeys), can now be used from inside ChatGPT, Claude, Microsoft Copilot, Gemini for business and Databricks Genie. In ChatGPT it is installed as a plugin. The announcement says Amazon Quick, IBM and Snowflake will come soon, although Adobe's documentation already walks through setting it up in Amazon Quick and in IBM watsonx Orchestrate.
The company made the agent generally available in June, as The Next Web points out. With this integration, marketing teams will be able to plan, coordinate and improve their customers' experiences from the AI assistant they already keep open. Among its functions Adobe lists analyzing how campaigns perform, forecasting audiences and optimizing customer journeys, always with people overseeing it. The agent rests on open standards such as Model Context Protocol (MCP) and Agent2Agent (A2A).
Adobe also introduced connectors with ready-made MCP servers to link the agent to other companies' systems: among them are Amazon Ads, Demandbase, Microsoft Work IQ, RainFocus, Snowflake Cortex AI and a connector for Meta campaigns.
The requirements are in the documentation: the agent has to be enabled for the organization and each user needs their own access in Adobe, plus Adobe Experience Platform and, at minimum, one of three products: Real-Time CDP, Journey Optimizer or Customer Journey Analytics. On several platforms, an administrator has to approve or enable it first. Inside them, the agent offers read-only analysis, does not change customer data and follows the permissions each person already has in Adobe. The announcement does not make clear whether each integration is generally available or in a testing phase.
Why it matters · analysis and opinion
What changes is the way in: Adobe's agent no longer requires opening its applications and instead sits in the assistant the marketing team has at hand all day. For Adobe it is a way of not being left out while companies choose their general AI provider. The door, however, only opens for those who already pay for its customer data platform, so it is not an entry route for small companies that do not have it. Inside those assistants, moreover, the agent returns read-only analysis, so its role there is to inform rather than to execute. Before turning it on, it is worth having the administrator review what permissions each user has in Adobe, because those are the ones the agent inherits.
Official source: Adobe · Written with the help of AI: how we make the news
