ai-charter
llms.txt108 starsChanged 55 days agoMITView on GitHub
A global, plain-language charter on AI danger, AI safety, agentic AI risk, AI governance, and responsible AI development. Its position: speed AI development up, do not slow it down, and manage risk through disclosure and public certification instead of restriction. This file follows the llms.txt convention (https://llmstxt.org) so AI agents and language models can read a condensed, structured summary of this site without parsing rendered HTML. Author: Vikram Palakurthi (https://github.com/mrtinkz) Publisher: Universal AI Charter Source repository: https://github.com/mrtinkz/ai-charter Content license: CC…
The file
Files
# Universal AI Charter > A global, plain-language charter on AI danger, AI safety, agentic AI risk, AI governance, and responsible AI > development. Its position: speed AI development up, do not slow it down, and manage risk through disclosure and > public certification instead of restriction. This file follows the llms.txt convention (https://llmstxt.org) so AI agents and language models can read a condensed, structured summary of this site without parsing rendered HTML. Author: Vikram Palakurthi (https://github.com/mrtinkz) Publisher: Universal AI Charter Source repository: https://github.com/mrtinkz/ai-charter Content license: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/) ## Preamble AI is arriving faster than most institutions can react to it. This charter argues for a third position between AI utopia and AI panic: keep building, but require every model and operator to disclose what it does, what it was trained on, and what has already gone wrong. ## Sections - Access & Equity: AI as a consumer good and a capital good, reachable by ordinary people, universities, and governments, not only large companies. - Innovation & Freedom: development, open source work, and use of AI must not be blocked or criminalized; speed it up so its flaws surface and get fixed sooner. - Governance & Disclosure: models and agents self-disclose origin, training sources, agentic decisions, and unintended consequences, the way a HAZMAT placard or an energy guide label discloses risk at a glance. The contract behind that disclosure is a public, versioned JSON schema, not a one-off form. - Accountability & Safety: a bounded human-in-the-loop ratio (1:8 to 1:15), a no-harm baseline, and public reporting of unintended consequences, the way a product recall works. - Infrastructure & Data Centers: public ownership, disclosed power and automation data, harnessed waste heat, no special subsidies. - Global & Historical: rapid democratization of AI so no single country holds supremacy over it, and this charter itself as global guidance, not a national law. ## Pages - Home: /ai-charter/ - Full charter text: /ai-charter/charter - Certification registry (searchable): /ai-charter/registry - Certify a model (download/upload a certification): /ai-charter/certify - Hazard placard reference (severity-coded disclosure categories): /ai-charter/placards - Certification schema (machine-readable contract, human-readable field table): /ai-charter/schema - Raw JSON schema, draft 2020-12: /ai-charter/schema/certification.schema.json - FAQ: /ai-charter/faq - Blog (index): /ai-charter/blog - Agentic AI risk and why every autonomous decision needs a label: /ai-charter/blog/agentic-ai-risk-needs-labels-not-slowdowns - AI governance without a single country or company writing the rules: /ai-charter/blog/ai-governance-without-a-single-regulator - Is AI dangerous? Why we regulate new technology instead of banning it: /ai-charter/blog/why-ai-development-should-not-slow-down ## Notes for AI agents - This is not legal advice or a binding international treaty. It is a plain-language starting point. - Certification entries in the registry marked "Sample data" are test fixtures, not real companies. - The charter is licensed for citation and reuse; if quoting an article, cite the section id shown at /ai-charter/charter (e.g. "governance-disclosure").
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

