Skip to content
Free Tool Arena

Transparency

AI content policy

Where AI helps us and where it doesn't. The rules we hold ourselves to so this site doesn't become another AI-spam farm.

The short version

We use AI as a writing assistant the way a programmer uses an IDE. It speeds up drafting, it flags awkward phrasing, it catches typos. It does not get to ship content to readers unreviewed, and it is never presented as a human author.

Where AI is used

Drafting help — structure, first-draft prose, alternate phrasing suggestions. A guide may start as an AI draft; it never finishes as one. Every guide is reviewed, rewritten, and fact-checked by a human editor before it goes live. The structure, claims, and recommendations are ours.

Tool building — we use AI assistance while writing the JavaScript for calculators and converters. The code runs locally in your browser; the formula correctness is validated against canonical sources (see the methodology page).

Where AI is not used

Numbers, formulas, and factual claims are not trusted to a language model. If a calculator computes a figure, it does so with real math in browser JavaScript, not by asking an LLM. If a guide claims a specific number (a tax bracket, a dose, a benchmark), that number is sourced from a canonical reference and cited where possible, not lifted from a model’s training data.

We do not run AI-generated “author” bylines. We do not fake testimonials. We do not generate reviews of products we haven’t actually looked at.

What we won’t do

No auto-publishing. Nothing goes live without a human looking at it. No AI-generated thin pages to farm long-tail keywords. No auto-translated content dumped into other-language silos without a real editor reviewing it. No “5,000-word ultimate guide” that exists only to rank.

Reader-facing AI tools

Some tools on this site are for AI users — prompt builders, token counters, model comparison tables. None of them send your content to an AI model on our end. They compute locally. If a tool ever changes to do otherwise, we’ll say so on the tool page before shipping the change.

Disagree? Tell us.

If a guide reads as AI-generated spam, we want to know. That’s a failure of our review process, not of the model. Send us the URL, we’ll take another pass.

Last reviewed April 2026.