Trust
AI usage policy
We build an AI-assisted writing tool, and we use AI in our own publishing. This page describes how — what we do, what we do not do, and what readers and users can expect.
In the product
RewriteAIForMe uses large language models to power its rewriting, tone, and humanisation features. We work with leading model providers under zero-retention terms: prompts and rewrites sent to underlying models are not retained by those providers and are not used to train their models. The rewrites you generate belong to you. We do not use the substance of your writing to train any model — our own or a third party's — under any plan.
What we store
We store your account-side rewrite history so you can revisit and refine prior work. You can delete any rewrite individually from your history, or request full account deletion from your account settings. We process deletion requests within thirty days. Aggregated, fully anonymised structural signals (such as sentence-length distributions across the platform) may inform our product and our research, as described in our research methodology; no identifiable content is included.
In our publishing
Our articles, guides, glossary entries, and research notes are produced by our editorial team. Where AI assistance is used — typically for structural drafting, formatting, or as an editorial sparring partner — a human editor is responsible for the published work: its argument, accuracy, voice, and the byline. We do not publish unedited AI output. We hold the same standard we ask of our customers: AI may help; the byline is still earned. Our full editorial principles are documented on the Editorial Standards page.
What we don't do
We do not generate large volumes of thin pages to capture search traffic. We do not fabricate statistics, citations, or quotes. We do not use AI to impersonate named individuals. We do not produce content with the primary purpose of manipulating rankings. We do not use customer content to train models. We do not provide tools designed primarily to evade detection — our work is designed to make writing genuinely better, with detector outcomes as a downstream consequence.
Model providers and updates
We periodically change the underlying models that power the product as the field advances. Changes that materially affect quality, behaviour, or pricing are communicated. We maintain the right to switch providers to deliver better output, lower latency, or improved privacy posture; we do not maintain the right to compromise the data terms described above.
Questions about how we use AI?
We are happy to answer. Get in touch with anything that affects whether you can use RewriteAIForMe with confidence.