ActiveCampaign Opens Its AI to Stripe, Webflow, Google Ads and Any MCP Server

ActiveCampaign has opened its AI layer to outside tools. On 12 November 2026 the marketing automation platform announced that Active Intelligence, its built-in AI assistant, can now connect to prebuilt connectors for apps such as Stripe, Calendly, Webflow, Google Ads and LinkedIn Ads, and to any app that exposes a Model Context Protocol (MCP) server.

The idea is that the AI drafting your campaigns can read data that lives outside ActiveCampaign, such as bookings, payments, website content or inventory, and act on it inside the account. ActiveCampaign says MCP connections are available now on all of its plans.

What ActiveCampaign announced

ActiveCampaign already exposes its own MCP server, so that assistants such as ChatGPT and Claude can read results, update contacts and build campaigns inside an ActiveCampaign account. This launch goes the other way. On its AI and MCP page, the company now describes two directions:

• Inbound: ActiveCampaign inside ChatGPT or Claude, connected through the ActiveCampaign MCP server URL found under Settings, Developer.
• Outbound: other apps inside Active Intelligence, connected by switching on a prebuilt connector or pasting the MCP server link of any other app.

Setup for the outbound side takes three steps, according to ActiveCampaign: switch on the connector or paste the MCP link, sign in with your own account for that app, then ask Active Intelligence to use it. The company says the MCP support builds on Active Intelligence 3.0, named Wavelength, which uses account history and past performance to suggest and act on campaign opportunities.

Chai Atreya, ActiveCampaign's Chief Product and Technology Officer, framed the launch around data marketers already hold: "Marketers already have the information they need to run great campaigns." Webflow's Chief Product Officer, Ben Haefele, is quoted in the same release: "Agents are only as good as the context they have."

What the connections are meant to do

The release gives four examples of campaigns built from outside data:

• Custom business data: a fitness studio pulls attendance records from its booking system to find members who stopped coming and build a win-back campaign.
• Paid advertising: marketers build Google Ads and LinkedIn Ads campaigns through a guided flow; Active Intelligence drafts the creative from existing messaging and sets up the campaign in the connected ad account, and the marketer reviews, approves and publishes it.
• Website to email: a travel agency combines Webflow destination guides with travel preferences collected in forms to build a targeted email campaign.
• Inventory: a specialty retailer turns overstocked products into a clearance campaign.

The press release headline also names Airtable, which does not appear in the list of prebuilt connectors in the body of the release; tools that are not on that list would be added through their own MCP server.

Permissions, approvals and AI credits

ActiveCampaign puts the access model at the centre of the launch. Each person connects with their own account in the connected app, so Active Intelligence can only reach the systems and data that this user is already allowed to use. Agencies can connect each client's apps under that client account's permissions. Ads are not published until a person approves them.

On cost, the AI and MCP page says connecting an app does not use AI credits; credits are spent when Active Intelligence builds or runs something, such as a campaign or a scheduled task. For company ChatGPT or Claude workspaces, the inbound connection to ActiveCampaign needs an administrator to set it up and approve it.

Where this fits

Email platforms have spent 2026 shipping MCP servers so that outside AI assistants can drive them. SMTPedia has covered launches from Iterable and Selzy. ActiveCampaign's move adds the reverse direction: the platform's own assistant becomes an MCP client that reads other tools. For senders, that means more of the data behind a campaign can come from systems outside the email platform, and more of the targeting logic is set up by the AI on request.

Action checklist

• List the data you would connect and check that each source is covered by your consent and privacy notices before an AI uses it to target email.
• Connect with the right account: since the AI inherits the permissions of the person who signs in, avoid connecting apps with admin accounts that see more data than the campaign needs.
• Review every segment the AI builds before sending, especially win-back and re-engagement audiences that can include old or inactive addresses.
• Verify older lists pulled from outside systems before mailing them, to keep bounce rates under the limits set by Gmail, Yahoo and Outlook.
• Vet custom MCP servers before pasting their link: only connect servers you or a trusted vendor operate.
• Track AI credit use once campaigns and scheduled tasks start running through Active Intelligence.

Sources

1. ActiveCampaign press release: ActiveCampaign Lets Marketers Bring Google Ads, LinkedIn, Webflow, Airtable, and Any Tool They Use Into Active Intelligence, 12 November 2026 (primary source)
2. ActiveCampaign: AI and MCP

Related on SMTPedia

• ActiveCampaign review: pricing, features and deliverability
• Iterable launches a hosted MCP server for Claude, ChatGPT and Cursor
• How to verify email addresses in ChatGPT and Claude with MCP