Adform has launched a Model Context Protocol server that allows advertisers to connect AI agents directly to its FLOW platform. The integration enables automated media planning, creative optimization, and reporting by connecting AI tools like ChatGPT and Claude to Adform's DSP and Ad Server.
The shift toward agentic workflows suggests a move away from manual ad operations toward high-level strategy. By allowing AI agents to interface directly with the backend of a DSP and Ad Server, Adform reduces the friction of complex media planning and data entry errors that plague traditional UIs. For the streaming and digital ecosystem, this signals a transition where technical proficiency in specific platforms becomes less critical than the ability to direct AI agents toward business outcomes like incrementality. Watch for whether competitors like The Trade Desk or Google launch similar open protocol servers to prevent losing agency traders to more automated environments.
Adform has launched an agentic advertising platform that connects external AI tools to its FLOW DSP and Ad Server using the Model Context Protocol. This integration allows AI agents to automate 800 platform capabilities, including media planning and reporting, reducing manual errors and shifting ad operations toward high-level strategic management.
Adform now allows brands to connect AI tools like ChatGPT, Claude, and Microsoft Copilot directly to its FLOW platform using a Model Context Protocol server.
The AI agents can access and automate 800 platform capabilities within the Adform FLOW DSP and Ad Server.
Dell Technologies and Publicis Groupe participated in early testing, focusing on reducing manual errors in taxonomy and creative management.
It reduces friction in complex media planning and data entry by allowing AI agents to execute repeatable tasks directly through the backend, rather than requiring manual UI navigation.
Add StreamingMeme as a preferred source on Google to see more of our streaming news at the top of your Search results.
Add as preferred source
