Set up AI access for your team
Learn how to set up Webflow AI across your workspace and roles. This guide shows how to control access and permissions so your team can use AI effectively without overstepping limits.
Video transcript
Your team most likely wants to use AI because that's what everyone wants these days. But before anyone starts prompting on your site, there is an important decision to make. Who should have access and what they should be able to do with it. In weblow, we'll set it up in two places. First, we'll do it at the workspace level and then with roles and permissions. First, let's go to our workspace, then general settings. And here is where we can enable Webflow AI across the entire workspace. That includes web flow provided AI features like AI assistant, MCP, AI site builder, AI code components, AI powered AO and more. If your plan includes those features, this one toggle controls whether they are available in your workspace. And a quick note, you need to be a workspace admin or owner to change this toggle. One more thing before we move on, if you are on a plan with Waflow optimizer localization, turning this off won't stop their core features. They use AI, but they're controlled separately. Okay, that's the workspace level setting. Now, let's look at how this works with roles. Let's go to team, then roles. Once the workspace toggle is on, default roles can use AI within their existing permissions. Let's open up a default role like marketer, and view the details. Here we can see use Webflow AI is enabled by default. But this is the important part. AI follows the same permissions the role already has. Turning on AI doesn't expand what someone can do, it expands how they can do it. So, a marketer with AI on can generate and refine copy and use AI in the templates and components they already have access to. But they still can't create or modify styles or components because that's not part of the role permissions with or without AI. A content editor with AI on can use AI for the work their role already allows like copy editing or editing CMS items, but they still don't get component editing or canvas building access. And a designer with AI on can use AI across the regular workflows, building, styling, creating components, design system changes. But again, AI is still limited to what the designer role can already do. So the rule is simple. Whatever someone can already do in the role, they can now do with AI. Whatever they couldn't do without AI, they still can't do with AI. Now, default roles cover a lot, but sometimes we need something more specific. That's where custom roles come in. Let's say we have a junior designer and we want them to keep designing manually, but we're not ready to give them AI access yet. In a custom role, we can start with the designer baseline role and deselect use Webflow AI. Now they can still do the other things the role allows. They just won't be able to use Webflow AI while doing them. Or the opposite. Let's say we have a marketer who needs to edit global components with AI. We can create a custom role based on the marketer role. Enable global component editing and keep use AI enabled. Now that marketer can edit components across the site using MCP AI assistant or manually. So that's how it works. First, toggle on AI at the workspace level so default roles can use AI within their existing role limits. Then, use custom roles when the default roles don't quite match what someone should be able to do with or without AI. AI can help your whole team move faster, and roles and permissions help ensure that everyone has the right level of AI access.
