How We Use AI
Readers deserve to know how a story was actually made. Here's exactly how AI is and isn't involved in what you read on NewsForge.
What AI does here
Articles on this site begin as reporting from other, credited news sources. An AI language model rewrites that reporting — restructuring it, adjusting tone and length, and adding context — rather than our staff writing it from scratch. The original source is always linked at the bottom of every article.
What AI does not do
- It does not invent facts, quotes, or events that weren't in the original reporting.
- It does not decide what counts as true — that's a human editorial judgment, not an AI one.
- It is not presented as a human reporter with a byline pretending to be a person.
Human review
When an article shows "Edited and approved by" a named person, a human editor on our team reviewed the AI-rewritten version against the original source before it was published. Articles without that line moved through automated publishing — we're not going to pretend otherwise, since that distinction matters for how much scrutiny you should give a piece.
When we get it wrong
AI rewriting can introduce errors — a misread number, a dropped nuance, an out-of-context phrase. When that happens, we post a dated, public correction directly on the article rather than quietly editing it. If an error is serious enough, we retract the story entirely and say so clearly at the top of the page.
Why we're doing it this way
We'd rather be honest about using AI and earn trust through transparency than pretend a small outlet has a large traditional newsroom it doesn't. Readers can decide for themselves what that's worth — but they should get to make that decision with accurate information, not a vague description designed to obscure it.