PropellerAds is bringing more of the media-buying workflow into conversational AI, making its advertising platform accessible directly through Claude via a Model Context Protocol (MCP) connector now listed in Anthropic’s Connectors Directory.
The integration allows advertisers to interact with PropellerAds campaigns through natural-language instructions instead of repeatedly moving between advertising dashboards. It reflects a broader shift in AdTech toward AI systems that can not only analyze campaign data but also execute operational tasks on behalf of media buyers.
The PropellerAds connector provides 39 tools spanning six functional groups, including campaign management, targeting, creative operations, bidding insights, analytics, account information and reference data.
From chatbot assistance to campaign execution
MCP is becoming an important piece of infrastructure in the emerging agentic AI ecosystem. Rather than asking an AI model to generate advice based on manually supplied information, an MCP connector can give an AI application structured access to external tools and services.
Anthropic describes MCP connectors as a way for Claude to interact with external systems directly, while maintaining the underlying integration as a separate connection between the AI application and the provider’s system.
For advertisers, the practical difference is significant.
With PropellerAds connected to Claude, a media buyer can create and modify campaigns, start or pause campaigns, change target URLs, adjust targeting parameters, review recommended bids, manage creatives and inspect performance without opening the conventional PropellerAds interface for every task.
The connector supports Interactive Ads, Telegram Ads, Paid Social, Onclick and Push formats. Targeting can also be adjusted across parameters such as geography, device, operating system, language, zone and sub-zone.
That does not turn Claude into an autonomous media buyer by itself. The advertiser still initiates the action. But it moves the interface for campaign operations from a dashboard toward a conversational agent.
Why MCP matters to AdTech
The announcement comes as advertising platforms experiment with a new layer of AI-driven campaign execution.
The Interactive Advertising Bureau’s 2026 outlook found that two-thirds of advertising buyers are focused on agentic AI for ad buying and campaign execution, while 96% are aware of agentic AI-powered buying. The same study forecasts U.S. advertising spend will grow 9.5% in 2026.
The significance of PropellerAds’ integration, therefore, extends beyond one advertising platform. It illustrates how MCP can become a bridge between general-purpose AI assistants and specialized AdTech infrastructure.
A media buyer could theoretically ask an agent to identify underperforming campaigns, examine available performance data and then make a specific change through an authorized API connection. The technology reduces the number of manual steps required to execute routine operations.
That is different from generative AI used solely for copywriting, brainstorming or reporting. Here, the AI sits closer to the execution layer of advertising.
IAB has identified this movement as part of the broader emergence of agentic media execution, where AI systems increasingly participate in planning, optimization and campaign operations under human oversight and defined guardrails.
PropellerAds is not alone in the race
The competitive landscape is evolving quickly.
DSPs, ad networks and other media platforms are experimenting with AI-powered campaign assistants, automated optimization and natural-language interfaces. PropellerAds itself has been developing NIKO, an AI agent designed to create, edit and optimize campaigns across its platform.
The MCP approach takes the idea in a somewhat different direction. Instead of requiring advertisers to adopt a proprietary AI interface, the connector lets them use an MCP-compatible agent such as Claude to interact with PropellerAds.
That distinction could become important as enterprises accumulate multiple AI assistants and specialized advertising systems. If connectors become standardized, advertisers could potentially interact with different parts of their marketing stack through a common AI interface rather than learning a separate automation layer for every platform.
The challenge will be maintaining control.
An advertising agent that can change bids, targeting or campaign status has access to functions that can directly affect media spend. Enterprises will therefore need clear permissions, authentication controls, audit trails and approval workflows before conversational campaign management can be trusted for larger budgets.
API tokens replace dashboard credentials
PropellerAds says its connector uses a dedicated API token rather than requiring advertisers to provide their account username or password. The token can be revoked independently, while campaign changes occur only when explicitly requested by the user.
That architecture provides an important distinction between AI access and autonomous authority.
The agent can perform actions, but it does not receive unrestricted permission to independently decide what to change. This human-in-the-loop model is likely to remain important as AdTech platforms give AI increasingly direct access to bidding, targeting and optimization systems.
It also reflects a broader industry concern: AI systems are becoming capable of acting inside advertising infrastructure, but accountability still needs to remain with advertisers and their technology partners.
The next AdTech interface may be conversational
PropellerAds’ Claude connector points toward a larger change in how advertising technology could be operated.
For decades, media buying has relied heavily on dashboards containing campaign tables, filters, bidding controls and reporting interfaces. AI agents introduce another possibility: advertisers describe the desired outcome conversationally while the system translates those instructions into API operations.
The immediate benefit is operational efficiency. Media buyers can spend less time navigating interfaces and more time evaluating strategy, creative performance, audience quality and budget allocation.
But the longer-term opportunity is more consequential. If MCP and similar standards become widely adopted, AI could evolve into an orchestration layer sitting above DSPs, ad networks, analytics platforms and other advertising infrastructure.
That would not eliminate media buyers. Instead, it could change their role from manually operating campaign interfaces toward supervising AI-assisted execution, validating recommendations and setting business constraints.
For AdTech vendors, that creates a new competitive requirement: having strong campaign automation may no longer be enough. Platforms will increasingly need APIs, permission models and interoperable AI interfaces that allow their capabilities to be accessed safely by the agents advertisers already use.
Market Landscape
The PropellerAds announcement lands in a market where agentic AI is moving from experimentation toward operational advertising workflows. IAB’s 2026 research says 96% of buyers are aware of agentic AI ad buying, while two-thirds are already live, testing or planning adoption.
The industry is also moving beyond AI as a reporting or recommendation tool. IAB says AI is increasingly influencing planning, bidding, optimization, pricing and measurement across the programmatic ecosystem.
This creates three important competitive battlegrounds:
- Natural-language campaign control: Letting buyers interact with advertising infrastructure conversationally.
- Agent interoperability: Connecting AI assistants to DSPs, ad networks, analytics and measurement systems.
- Governed execution: Giving agents enough permission to act while retaining human approval, auditability and spending controls.
The strategic question for enterprise advertisers will be whether these systems merely make existing workflows faster or eventually change how media buying itself is organized.
Top Insights
- PropellerAds’ MCP connector gives Claude users conversational access to campaign creation, targeting, bidding, creatives and analytics across five advertising formats.
- The integration moves generative AI closer to actual campaign execution, rather than limiting AI to recommendations, reporting or content generation.
- IAB data shows agentic AI has become a major advertising priority as buyers seek greater automation across planning, buying and optimization.
- API-based authentication and explicit user commands provide governance mechanisms as AI agents gain access to functions that can affect media spending.
- MCP could become an important interoperability layer connecting general-purpose AI assistants with specialized DSP, ad network and advertising technology infrastructure.
Get in touch with our Adtech experts
