On July 1, TikTok did something the other major ad platforms have also been doing: it opened its ad platform to AI agents.
Agentic Hub launched alongside TikTok's MCP server — two pieces of infrastructure that, together, let any MCP-compatible AI assistant connect directly to your TikTok Ads account and take action inside it. Create campaigns, pull performance data, adjust targeting, manage creative operations. No login required, no dashboard navigation, no manual export.
If you use Claude, ChatGPT, or any other AI tool that supports MCP, you can connect it to TikTok Ads today.
What the MCP server actually does
MCP (Model Context Protocol) is an open standard that lets AI assistants connect to external services and take actions through them. When TikTok built an MCP server for its ads platform, it gave AI agents structured access to everything inside TikTok Ads Manager.
The scope covers campaign management, performance reporting, audience configuration, and creative operations. An agent with this access can query your account, flag underperforming ad groups, pull a breakdown of spend by creative, and surface recommendations — or execute changes you instruct it to make.
For accounts running GMV Max campaigns, which TikTok confirmed is now the default campaign type for all TikTok Shop ads, the data volume that benefits from automated monitoring is significant. Creator affiliate performance, product catalog coverage, bid pacing, shop-level sales attribution — an agent running regular checks on all of that has a use case that's easy to see.
Agentic Hub: the Skills marketplace
The MCP server is the connection layer. Agentic Hub is the discovery layer — a marketplace where you find pre-built AI Skills built on top of that access.
Fourteen partners launched Skills at the July 1 rollout. The categories span different advertiser needs.
HubSpot and Constant Contact connect your CRM and email data to campaign decisions. Wix integrates storefront activity so your agent has context about site behavior alongside ad performance. Kochava brings attribution modeling. Innovid handles creative measurement and analytics.
On the ecommerce infrastructure side, Shoplazza connects TikTok Shop directly to your store inventory, so an agent can match catalog health to campaign spend in one view.
Agency and performance-network partners — Madhouse, Mobvista, HuntMobi, Cyberklick, Storyverse, Bellnova, and AI Rudder — have each published Skills for specific regional markets or campaign workflows.
TikTok also supports custom Skills. If you or your agency want to build something specific to your account structure — a GMV Max monitor, a product-tier bidding agent, an approval flow for new creatives — the MCP server is the foundation, without requiring a custom API integration.
The pattern across platforms
TikTok isn't alone here. Meta launched its MCP server for Ads Manager in May, giving AI agents the same kind of access to Meta campaigns. Google followed with its own MCP integration around the same time. Snap has a beta implementation running.
Every major ad platform is building the same thing: a standardized access layer that lets AI agents manage campaigns directly. The manual, login-and-click interface is still there. It's just no longer the only option.
What's changing isn't just convenience. It's where campaign management work actually happens. Right now, most of that work lives in dashboards — you log in, pull reports, identify problems, make changes, repeat. As MCP access becomes standard across platforms, more of that cycle moves into AI-assisted workflows. You describe what you want, the agent handles execution.
The obvious caution: an agent with write access to your ad account can make changes at speed. The same supervision you apply to automated bidding applies here. You remain responsible for what runs in your account, and TikTok's terms are clear about advertiser responsibility for actions taken through connected tools. Start with read-only access — reporting and analysis — before granting write permissions, and understand what an agent is doing before you let it execute.
The starting point
You don't need to build a custom agent to get value from this. The Skills in Agentic Hub are ready to use. Browse the hub, identify which partners your team already works with, and evaluate the workflows each Skill enables.
The most immediate question: what are you doing in TikTok Ads Manager manually and repeatedly? Pulling weekly performance reports. Flagging creatives with declining CTR. Checking whether your catalog is fully covered by active ad groups. Those tasks are now automatable if you configure the right agent for them.
If you want to understand what's actually running in your TikTok account — what's performing, what's running on defaults, where your budget is actually going — the free account audit at Gromerce gives you a clear baseline in a few minutes.
Sources: TikTok for Business Blog, Digiday, MediaPost, Social Media Today, PYMNTS, July 2026
