Answers to common questions

A deliberation platform where multiple AI models — your "senators" — debate a question, and a Chief Senator you choose compiles their input into a single opinion you can review, vote on, and export.

Yes. Senate of AIs™ uses a Bring-Your-Own-Key model. Your keys are encrypted (AES-256-GCM) and are only used to call the providers you link. Free-tier providers such as Google Gemini, Groq, Mistral, and DeepSeek are labeled on the Senators page.

OpenAI, Anthropic, Google Gemini, xAI Grok, Groq, Mistral, DeepSeek, and Cohere. Each senator persona (Stoic, Populist, Analyst, and others) can be linked to any supported model.

You post a question, the linked senators deliberate in parallel, then the Chief Senator you selected compiles a single opinion. You review and edit that opinion before the senate votes on it.

After the Chief Senator compiles the opinion and you approve it. The vote is on the compiled opinion, not on the raw deliberation.

Yes. Every edit is saved as a version in the Opinion Timeline, and you can compare any two versions side-by-side with a color-coded diff.

Word (.docx) and PDF for opinions, version history, drafts, and diffs. All exports carry the Senate of AIs™ / Marcus Institute cover page, running headers and footers, and the Human-Judgment Disclaimer.

Yes. The Compare dialog produces a short shareable link that opens with the same Left and Right versions and view mode you chose.

Senate of AIs™ and Chief Senator™ are trademarks of Keith David Whalen, published under The Marcus Institute. See the Trust & Safety page for the full framework.

The platform blocks and reports content covered by the Trust & Safety framework — self-harm, violence, illegal activity, exploitation, and harassment. Use the in-app report button on any session.

No. Every opinion carries a Human-Judgment Disclaimer. Senate of AIs™ is a decision-support tool, not a replacement for legal, medical, financial, or other professional advice.