How to Add an AI Chatbot to Google Sites

The practical way to add a chatbot to a Google Sites page is to paste a JavaScript embed snippet from a third-party chat service into the page's Embed block. Google Sites does not ship its own AI chat feature, so you are relying on an external widget. Tools like fenbs can keep the approved FAQ sources and change requests your team agrees on, so the next person to update the chatbot starts from a clear record rather than an old chat thread.

What Google Sites allows

Google Sites supports embedding external content through the Insert > Embed menu. You paste a URL or an HTML snippet and the site renders it in an iframe. The catch is that Google Sites applies a content security policy that blocks some types of scripts. Not every chat widget will render correctly; a widget that works fine on a standard web page may be blocked or silently fail in a Google Sites iframe.

Before picking a service, test its embed snippet on a published Google Sites page, not just a preview. Published pages apply the real security policy; previews sometimes behave differently. If the widget does not appear, check the browser console for content security policy errors.

Option 1: Embed a third-party chat widget

Services such as Tidio, Crisp, and similar tools provide an embed snippet you paste into a website. In Google Sites, go to Insert > Embed, choose "Embed code," and paste the snippet. Resize the embed block to match the widget's expected dimensions. After publishing the page, test the widget in a real browser to confirm it loads and can receive input.

1. Create an account with your chosen chat service and set up the chatbot, including its name, greeting, and knowledge sources.
2. Copy the embed code from the service's widget or installation settings.
3. In Google Sites, open the page you want to add the chatbot to, go to Insert > Embed > Embed code, and paste the snippet.
4. Resize and position the embed block. Click Publish to make the change live.
5. Open the published URL in a fresh browser window and test the chatbot. Check that it loads, accepts a message, and responds correctly.

If the widget does not appear after publishing, open the browser developer tools and look at the Console tab for content security policy errors. Some services offer a separate iframe URL option that is more compatible with restrictive embedders; check the service's help documentation.

Option 2: Embed a public Dialogflow CX agent

Google's Dialogflow CX lets you publish a conversational agent and generate a test or messenger URL. You can embed the messenger URL as an iframe snippet in Google Sites. This approach stays within the Google ecosystem, which can simplify sign-in and access. The tradeoff is that Dialogflow CX requires more setup effort to configure intents, training phrases, and connected data sources compared to a simpler out-of-the-box chat widget.

What to check about data and privacy

When a visitor types a question into a chat widget embedded on your site, that message is sent to the widget service's servers. If your Google Sites page is internal, such as a company intranet or a team project site, confirm that the chat service's data processing agreement is acceptable to your organization before you publish. Ask specifically where messages are stored and for how long.

For public-facing pages, include a brief note near the chatbot that explains how messages are handled. Visitors on a US-based site may have rights under state privacy laws. Check with your legal team if the site collects names, emails, or other identifiable information through the chat interface.

Keep your FAQ sources and change requests organized

A chatbot is only as useful as the knowledge it can draw on. When you build a chatbot for an internal wiki or a product support page, the approved sources, the tone guidelines, and the decision to include or exclude certain topics should live in writing, not just in someone's memory. When a colleague updates the chatbot six months later, they need to know which sources were approved and what was left out on purpose.

How to do this with fenbs

1. Create a free account and add a project for your Google Sites chatbot.
2. Record the approved knowledge sources as Decisions and rules, including any documents, URLs, or FAQs the chatbot is allowed to answer from.
3. Add a Links and listings entry for each connected service or widget you are using, including the service name and configuration location.
4. When a content owner requests a change to the FAQ, add it as a task in To Do so there is a clear record of what is pending before it is published.
5. Connect a supported AI assistant from Integrations to help draft chatbot responses or review the FAQ content, with your project rules in place.

fenbs does not connect directly to Google Sites or to your chat widget. It keeps the team's decisions about what the chatbot should say, and the change requests that are pending, in a record that survives a personnel change or a long gap between updates.