Can you use Adobe's marketing agent in ChatGPT or Claude?
Yes. On November 4, 2026, Adobe made CX Enterprise Coworker, its agentic AI for customer experience orchestration, available inside ChatGPT, Claude, Gemini for business, Microsoft Copilot and Databricks Genie. The bigger lesson for marketing leaders: the AI interface is becoming a commodity, and...
Yes. On November 4, 2026, Adobe made CX Enterprise Coworker, its agentic AI for customer experience orchestration, available inside ChatGPT, Claude, Gemini for business, Microsoft Copilot and Databricks Genie. The bigger lesson for marketing leaders: the AI interface is becoming a commodity, and the advantage moves to the data and rules behind it.
What did Adobe announce?
On November 4, 2026, Adobe announced that CX Enterprise Coworker is available within ChatGPT (as a plugin), Claude, Databricks Genie, Gemini for business and Microsoft Copilot, with Amazon Quick, IBM and Snowflake expected soon. The agent is built on Model Context Protocol (MCP) and Agent2Agent (A2A). Adobe also launched connectors with prebuilt MCP integrations to partners including Amazon Ads, Meta campaigns, Demandbase and Snowflake Cortex AI. Systems integrators such as Accenture and Deloitte Digital, and agencies including dentsu, Omnicom, Publicis and WPP, are building on it. MarTech listed it among the week's top releases.
My take: your Marketing OS is the edge
AI apps will be commoditized. Anyone can build one already. An app is just a window where someone types a request and expects a useful result. Its real job is to act as a guardrail, so the AI chat and each person's prompts don't turn into slop and the output is reliably good. That is valuable, but it won't stay rare, and Adobe putting its agent inside every major assistant shows that even the biggest vendors know the window isn't the moat.
What can't be copied is the backend the app sits on. I call it a Marketing OS: the well-designed foundation everything runs on. It has to be secure, and your most important data and business rules have to be defined in it: customers, products, brand standards, approval logic. That belongs to your company as an asset, and it is your company's edge. I made a similar case for structured data in Airtable for Marketers.
Once that foundation exists, MCP changes the economics. Make your Marketing OS available through MCP, and employees can connect the AI chat they prefer and do their work on it, without each tool rebuilding your context from scratch. One well-governed foundation can then serve every team and every tool.
What should marketing leaders do now?
- List the data and business rules your marketing AI depends on, such as customer data, product information, brand guidelines and approval rules, and name an owner for each. That list is the start of your Marketing OS.
- Secure it, then expose it through MCP with clear permissions, so teams can work from their preferred AI chat instead of pasting context into prompts.
- Before buying another AI app, ask whether it adds real guardrails on top of your foundation or just another window on top of someone else's data.
Get the AI Marketing Brief twice a week. Subscribe free for my take on the AI news that matters to marketing leaders.
Frequently asked questions
Which AI platforms support Adobe CX Enterprise Coworker?
As of November 4, 2026, Adobe CX Enterprise Coworker works inside ChatGPT (as a plugin), Claude, Databricks Genie, Gemini for business and Microsoft Copilot. Adobe says availability in Amazon Quick, IBM and Snowflake is expected soon.
What is MCP in marketing?
MCP, or Model Context Protocol, is an open standard that lets AI assistants connect to outside tools and data. For marketers, it lets an assistant like ChatGPT or Claude work with systems such as a CRM, CDP or ad platform, under permissions the company sets.
Sources
ZoomMetrix Newsletter
Join the newsletter to receive the latest updates in your inbox.