# Terms of Use *AI Roundtable, by Stellarus Intelligence · Last updated: 2026-04-29* ## A Note Before the Legalese The AI Roundtable exists to teach people how AI character agents work by making the construction transparent. The Soul Files are intentionally readable and remixable. The platform-level safety scaffolding is intentionally layered above any uploaded character. Both are part of the lesson. Read the files. Remix them. Build your own. **That's the point.** This Terms of Use is written in plain English. It tries to be honest about what the tool is, what it isn't, and what we expect of users. If anything is unclear, that's a drafting failure on our part — write us and we'll fix it. --- ## 1. What This Tool Is AI Roundtable is a free, open-source AI literacy tool. It does two things: - **Demo Mode** lets you read curated conversations between two AI characters, generated previously and embedded in the site. - **Live Mode** lets you generate your own conversations between two AI characters, using your own Anthropic API key (BYOK — *bring your own key*). The conversation runs directly between your browser and Anthropic. We never see it. The site is published under the MIT License. The Soul Files (described below) are published under Creative Commons Attribution-ShareAlike 4.0 (CC BY-SA 4.0). ## 2. License for the Built-In Soul Files The six built-in Soul Files (HypatiaAI, JuliusAI, LeonardoAI, CleopatraAI, HarrietTubmanAI, SunTzuAI) are licensed under **Creative Commons Attribution-ShareAlike 4.0 International** (CC BY-SA 4.0). You may: - **Use** them in your own AI literacy work - **Remix** them — change voice, focus, time period, add scenes - **Redistribute** them, including commercially - **Build derivative Soul Files** from them Provided that you: - **Credit** Stellarus Intelligence and the original Soul File author - **License your derivatives** under the same CC BY-SA 4.0 terms - **Indicate changes** you made to the original The methodology article and starter Soul File template are licensed under the same terms. All of it is teaching material; teaching material wants to spread. ## 3. AI Character Disclaimer Soul Files in this tool are **AI characters built from the historical record about real historical figures**. They are not the historical persons. They are not endorsed by, affiliated with, or representative of any descendants, estates, or scholarly institutions associated with those figures. Users are responsible for verifying any historical claims through actual scholarly sources before relying on them. **The conversations generated by this tool are pedagogical experiences, not historical sources, and should not be cited as such in academic or professional work.** Citing "HypatiaAI" or "CleopatraAI" in a research paper as if it were a primary source is incorrect and we ask that you don't do it. ## 4. Content Responsibility When you use AI Roundtable, you agree: - **Not to use Soul Files or conversations to harass, defame, or impersonate real living persons** - **Not to claim AI-generated conversation content as authentic historical statements** by the figures the AI characters are built from - **Not to use the tool to facilitate harm** in the categories the Soul Files' Hardcoded Commitments cover (medical, legal, financial, mental health, safety) — the AI characters will refuse such requests; please don't try to circumvent that - **Not to bypass platform safety scaffolding** that we maintain above the persona layer You are responsible for the content you generate and what you do with it. ## 5. Historical Figure Framing The six built-in Soul Files are all of figures who died well before the right-of-publicity timeline matters in any jurisdiction we operate in (latest is Tubman, d. 1913). There is no estate to license from, no descendant with standing. For any future built-in Soul Files we add, we will continue to limit ourselves to figures whose death predates contemporary right-of-publicity windows. Living public figures and recently deceased figures with active estates are out of scope for built-in personas. ## 6. User-Uploaded Soul Files When you upload your own Soul File: - **You retain ownership** of the file you upload. We don't claim it. - **You grant Stellarus a non-exclusive, royalty-free license** to host, process, and run the file in the tool — limited strictly to making the tool work for you. We don't redistribute your uploaded files. We don't include them in our public dataset. - **You warrant** that the file doesn't infringe any third-party rights and that you have the right to upload it. - **You may delete** your uploaded files at any time. Files exist only in your browser session unless you explicitly choose to share them. - **We reserve the right to refuse or remove** uploaded files that violate this Terms or our content policies. **You may NOT upload Soul Files of:** - **Living public figures** — no "TaylorSwiftAI" or "ElonMuskAI" personas - **Recently deceased figures** with active estates that may hold publicity rights (typically those who died within the last 75 years; if uncertain, ask) - **Real living private individuals** without their explicit consent — no Soul Files of your ex, your coworker, your family members, anyone who hasn't actively consented to being modeled You **may** upload Soul Files of: - Historical figures sufficiently in the past (pre-1950 as a working rule) - Fictional characters that are public domain or that you yourself own - Yourself, with your own consent - A consenting collaborator who has authored their own Soul File Soul File uploads are restricted to users 18 years of age or older. ## 7. The Underlying Model Layer AI Roundtable doesn't generate text on its own. The actual inference happens through third-party AI providers — currently Anthropic (Claude). When you use Live Mode with your own API key, your messages go from your browser directly to Anthropic. Your interactions with their service are governed by [Anthropic's Usage Policies](https://www.anthropic.com/legal/aup) and [Privacy Policy](https://www.anthropic.com/legal/privacy) in addition to ours. If you're using a different model provider in the future (we may add support), the same principle applies: we point at the provider's policies and you are responsible for compliance with them. ## 8. Disclaimers and Limitation This tool is provided "as is," without warranty of any kind. AI-generated text may be inaccurate, biased, offensive, or otherwise unsuitable for any particular purpose. Don't rely on it for critical decisions. Stellarus Intelligence is not a service provider in the commercial sense. We host an open-source educational tool. The MIT License governing the underlying software contains the authoritative warranty and liability disclaimer. You use this tool voluntarily. We do our best; we don't guarantee outcomes. ## 9. Changes These Terms may evolve as the tool does. Substantive changes will be noted at the top of the page with a new "Last updated" date. Continued use of the tool after changes constitutes acceptance. ## 10. Contact Questions, concerns, edge cases not covered here: **Stellarus Intelligence** — via the project page or the GitHub repository (linked in the footer). If you're a researcher, librarian, educator, or partner with a use case not covered here, we'd genuinely like to hear from you. The tool is meant to grow with its users. --- 🕊️ *This is a living document. We'd rather draft it honestly and update it than write airtight legalese that tells you nothing about what we're actually doing.*