Editorial policy and corrections

What we publish

Plain-English explainers about AI risks, the safe use of AI tools, AI governance and AI rules, for families, schools, small teams and people entering the field. Each page answers one question, shows its sources, and shows when it was last checked. We are a reading guide, not a news desk, so we publish explainers on a steady rhythm rather than breaking news.

What we will not publish

• Instructions for misusing AI: jailbreaks, ways around safety filters, ways to avoid AI detection, how to build deceptive deepfakes, or help with weapons or malware.
• Sexual deepfake content of any kind. Non-consensual intimate imagery appears only as a harm, with where to report it and get help.
• Legal, medical, financial or investment advice, stock picks about AI companies, or certification and compliance guarantees.
• Claims about private individuals, invented statistics, invented quotes or invented case studies.
• Content aimed at children under 13. Our family pages are written for parents.

How AI is used to make this site

We use AI tools to help research, draft and edit pages, and to make the photographs on the site. We do not hide that. Every page is built from linked sources and checked against them before it is published, and every image is labelled as AI-generated on the credits page. No page names a human author, because none is claimed: the site publishes as an organisation, and we do not invent expert personas or quotes.

Independence and funding

AI Safety Association is operated by Euclid Squared Inc. It is not a membership body, and no AI company, agency or advocacy group pays for or approves our pages. The site does not currently carry advertising or paid links. If that changes, we will say so on this page, label every sponsored or affiliate link where it appears, and keep it out of how we describe any product or rule.

Non-partisan

We describe rules, who proposed them and what they do. We do not tell you how to vote, and we do not praise or attack a party, a politician, a bill or a company. Where experts disagree, we say who disagrees and why.

Corrections

If you think something on the site is wrong, out of date or unclear, email [email protected] with the address of the page and, if you can, a link to the source that shows the problem. We read every message.

• We check the claim against the primary source. If we were wrong, we fix the page and move its last-checked date.
• A substantive fix (a changed fact, date or conclusion) is also noted on the page, so readers can see what changed. A typo or broken link is fixed without a note.
• If we cannot confirm the problem, we reply with what we checked and why the page stays as it is.

We aim to answer within five working days. Read how we check sources for the steps behind each page.