Name the concern
Point to the claim, safeguard, term, or process that needs work.
Public review brief
Common Ground Standard is asking a practical question: what should people be able to see, question, protect, and change before an AI system or infrastructure project is described as responsible?
Why review comes first
Students, workers, communities, public agencies, foundations, and institutions need ways to distinguish real public benefit from broad AI-for-good language. This draft is being opened early so its weak points can be found before broader foundation, agency, industry, infrastructure, or implementation outreach.
The work is not finished by design. A useful review makes the next version harder to misuse, easier to understand, and more accountable to the people and places affected.
What the draft is trying to protect
AI deployment should preserve human agency.
AI infrastructure should return visible community benefit.
AI-created gains should have a fair benefit path.
Data use should remain under clear, enforceable governance.
Ecological burdens should be visible, bounded, and repairable.
AI education should preserve student authorship and learning integrity.
Choose one place to start
Each page explains the subject in plain language, offers a five-way pressure test, and carries one focused question into the website feedback form.
What useful feedback looks like
Point to the claim, safeguard, term, or process that needs work.
Explain whose rights, work, learning, services, resources, or trust are at stake.
Suggest the evidence, authority, appeal, public record, or no-go condition required.
Recommend clearer language or a safeguard that can be checked in real use.
Where your feedback goes
The website form is the primary path and does not require a GitHub account. GitHub remains available for reviewers who want to discuss a public issue or propose a source edit. Longer institutional review can be requested through Contact.
This phase is public critique. It is not a funding request, partnership request, pilot request, certification process, petition, public sign-on, or request for endorsement. Do not submit confidential or restricted material.