AI Disclosure

Ted is AI

Ted is software powered by AI, not a person, even when talking with him feels natural. If you’re ever unsure in a conversation with Ted whether you’re talking with a person, you’re not: it’s Ted. (In support chat, people from our team may answer.)

What powers Ted

Ted runs on AI models from providers including OpenAI and Anthropic, with specialized services for things like images, web search and remembering earlier chats. The models we use change over time as better ones arrive. Our subprocessor list names every provider and what each one receives.

What Ted can do

• Read your messages and files, your foundable’s websites, and the apps and accounts you connect.
• Search the web and tell you where he found things.
• Plan work and carry it out: building and publishing websites, writing, making images and videos.
• Act in the services you connect, such as posting, sending email or texting you, within the permissions and settings you choose. If an owner or admin allows it, he can do this without asking again for each step.
• Remember earlier chats within the same foundable.

What Ted can’t do

• See your device, browser or accounts you haven’t connected.
• Guarantee he’s right. He can sound confident and still be wrong.
• Replace a lawyer, doctor, accountant or other professional.
• Take back something already done in another service. Stopping Ted stops future work, but a sent email can’t be unsent, and a published post can only be removed with that service’s own tools.
• Share what he knows between different foundables.

Your content and AI training

We don’t train AI models on your content. OpenAI and Anthropic, which power Ted, process it under API terms that don’t use it to train their models, and our subprocessor list says how each other provider handles it. Our Privacy Policy explains how we handle your information.

Check Ted’s work

Ted’s work can be wrong, out of date, biased, or not right for your audience, and it can resemble what he makes for others. He can even invent facts, names or sources. Review what he makes before you publish, send or rely on it, especially for important decisions.

If the law where you work requires you to say content was made with AI (for example, some political ads), that’s up to you. Some decisions about people need a qualified person’s review, and some uses aren’t allowed at all; our Acceptable Use Policy explains which.

Report a problem

• Harmful or unsafe output: abuse@foundable.com
• Wrong or biased output: support@foundable.com
• Security issues or ways around Ted’s safety rules: security@foundable.com

AI laws

This page explains how Foundable works. It doesn’t mean any particular use of it complies with AI laws such as the EU AI Act or US state AI laws; you’re responsible for the notices, reviews and labels your own use requires.