Using AI responsibly

Everything you create in Vibewax is built with AI. Some laws, including the EU AI Act, require people to be told when they are interacting with an AI system or looking at AI-generated content. For the projects you publish, Vibewax handles part of this for you and part of it stays your responsibility.

This page describes what Vibewax does for you and where your responsibility sits. It is not legal advice. Whether disclosing AI interactions or AI content applies to you depends on applicable law and how you use your project. If you are unsure, talk to a legal advisor.

What’s built in

In published projects

Vibewax adds a machine-readable <meta name="ai-generated" content="true"> tag to the site’s head when you publish. It tells crawlers, browsers, and other automated tools that the site was built with AI.

On generated media

Images, audio, and video produced through Vibewax carry whatever provenance markings the underlying provider applies, such as SynthID or C2PA. Vibewax does not strip these markings.

Do not intentionally remove or circumvent these markers where doing so would break the law that applies to you.

What’s in your hands

The platform’s own markings cover the platform. They say nothing about what your published project does, or what you do with its output. Those obligations stay with you.

AI features you add to your project

If your project includes something like a chatbot, content generation, AI summaries, or anything else powered by AI, your end users need to know they are communicating with AI. Vibewax does not add that notice for you, because it is part of your application rather than the platform itself.

You can ask the agent to add it, for example as a short note above the chat window.

Content you publish

The machine-readable tag says the site was built with AI. It says nothing about the content on it, and most visitors will never see it. If you publish text, images, or media that were generated by AI and your use case calls for a visible disclosure, adding it is up to you.

Synthetic media

Content that convincingly depicts real people, places, or events but is generated or manipulated by AI carries stricter obligations under the EU AI Act. If your project produces or publishes this kind of media, look into which rules apply to you before you ship.

Practical examples

I published a marketing site with AI-generated copy

The machine-readable tag records that the site was built with AI, but it is not a visible notice to your readers. Whether you need to disclose that the copy itself is AI-generated depends on applicable law and the specific use case.

I built a customer support chatbot into my app

Your end users need to know they are talking to AI rather than a person. Ask the agent to add a visible notice at the start of the conversation or above the chat window. The notice shown inside Vibewax does not carry over to your project.

My app generates images of people using AI

Images generated through Vibewax carry whatever provenance markings the provider applies. If the images could pass for real people or real events, additional transparency obligations may apply to you.

I'm building an internal tool that nobody outside my company uses

Transparency obligations still apply to the people using the tool, even when they are your own colleagues.

Reviewing your project with the agent

The agent can help you find the places where a disclosure might belong. Ask it in plain language:

Go through my project and list every feature that uses AI,
including anything that talks to an AI model. For each one,
tell me whether an end user would be able to tell it's AI.

Add a short notice above the chat window telling users they
are talking to an AI assistant, and make it match the rest
of the site's styling.

See Building secure apps for more on working with the agent.