
The Amazon Ads MCP Server, based on the open-standard Model
Context Protocol (MCP), is moving into open beta.
Paula Despins, vice president of ads measurement at Amazon Ads, made the announcement at IAB ALM during her keynote on Monday. She explained
how advertisers and partners can use one integration to link custom agents or AI platforms like Claude, ChatGPT or Gemini.
After setup, the Amazon Ads MCP Server allows these agents to use
Amazon Ads feature, including managing campaigns, checking performance reports, adjusting account settings, and viewing billing or financial data.
Technical processes are
streamlined reducing complex integrations. As Amazon's APIs evolve, AI agents remain compatible automatically, eliminating the need for code rewriting or frequent integration updates.
Advertising agents were introduced at Amazon Ad’s annual unBoxed event last year. Now the focus is on extending these agentic capabilities beyond its ads console.
advertisement
advertisement
This
connectivity alone does not guarantee reliable outcomes — especially in advertising, where real-world workflows often span multiple systems and decisions, according to Despins. Tasks such as
launching a campaign in a new country can involve dozens of manual steps.
“APIs remain essential for programmatic operations, but they were designed to expose specific
capabilities one at a time, not to coordinate the kind of complete, autonomous workflows that AI agents now need to manage,” Despins wrote in an accompanying blog post. “Without guidance,
agents must piece together how individual steps fit together, and they don't reliably produce the best results.”
The tools launching along with the open beta are intended to reduce
complexity by acting as an instruction manual, turning complex advertising operations into simple actions an agent can execute.
For users, tools are simple to invoke through natural language
prompts.
She also explained in the post how MCP tools reduce complex tasks for common advertising actions, so advertisers and partners can focus on unique use cases and what differentiates
them — their strategies, their experience, their expertise.
For example, if an advertiser is running campaigns in the U.S. and Canada, Amazon Ads has a tool that enables the advertiser
to quickly expand it to another country with one prompt.
Amazon Ads also launched a tool to create Sponsored Products campaigns, end to end, in one workflow.
The tool handles what would
normally require three or more operations — to create the campaign, such as setting up ad groups, and creating one prompt.
This generates an immediate campaign that only needs reviews
and approvals — helping advertisers launch campaigns faster while reducing manual setup.