AI Writing Disclosure

Plain English, and binding. A law firm did not write this page; the founder did, in plain English, so you can actually read it. That does not make it any less real: these are the operative terms of the Platform, and using the Platform means you agree to them. We will have counsel review these terms as the company grows, and any material change comes with at least 30 days notice, the same as our Terms of Use promise.

The short version: the Platform includes an AI writing assistant, and it is live today. It produces drafts on your blog. In the default setting nothing publishes until you approve it. There is a second setting you can switch on that publishes for you, and posts published that way say so, on the post, above the first paragraph. Every post the assistant writes carries a disclosure whether you approved it or not. You are the publisher of record either way. Raw machine output may not be protected by copyright, and the assistant will not write income claims for you or anyone. Here is exactly how it works.

Who Gets Drafts, and How Many

The assistant is part of the Blogging System subscription, and what you buy is an allowance, not a delivery schedule. An active subscription carries an allowance of 5 AI drafts a month; Blogging System Pro carries 25 a month. Without an active blog subscription you have none. You can add a pack of 10 drafts for $39 once (they never expire and are only drawn on after your monthly allowance runs out) or 10 extra every month for $35 a month. Unused monthly drafts do not roll over.

An allowance is not a promise that a draft appears. Generation depends on outside AI providers and on our own systems, and both can be unavailable, rate-limited or deliberately paused by us. When that happens you do not get drafts that month, and no draft is charged against your allowance for a request that failed. We are not going to dress that up: if the assistant is the reason you are subscribing, subscribe on the blog, the network and the guarantee, and treat the drafts as the part that is still young.

There is no gate on your published post count. Anyone with an active subscription can use the assistant from day one. An earlier version of this page described a three-post gate; that gate was never built, and leaving the claim here would have been a false statement about a live product.

Two Modes, and What Each One Does

Approval mode is the default, and it is what every account starts in. A draft the assistant produces is saved as a draft on your blog. It sits there. Nothing goes out until you open it and press approve. You can ask for a draft to be rewritten up to twice before you accept or discard it.

Auto mode is opt-in and you have to earn it. It only becomes available after you have completed at least one clean cycle of approving drafts yourself, and you have to switch it on deliberately. With auto mode on, any draft the assistant stages is published to your blog by the system, at a time it picks, without you reading it first. That is the whole point of the setting and it is the whole risk of it: posts have gone out on member blogs this way. You can switch back to approval mode at any time, and switching back does not un-publish what already went out. We can also hold an account in approval mode permanently — for example an account writing in a money-adjacent niche — and we will do that wherever we think a machine publishing unread is a risk to readers.

Two safety controls sit above both modes: a limit on how many generated drafts may be waiting on one blog at a time, and a network-wide switch that pauses all generation and all automatic publishing at once. We will use the second one without hesitation if the output stops being good enough to have our name on it.

Every Generated Post Says So

Plain English: readers are told, on the post, at the top. A post written by the assistant carries a disclosure line above the first paragraph, on the post page and in your RSS feed. It is generated from the post’s own record rather than pasted into the body, which means it cannot be edited out and it stays accurate if the post is revised later.

The wording differs by what actually happened, because those three things are genuinely different and a reader is owed the distinction:

• Drafted by the assistant, then reviewed and published by you.
• Drafted by the assistant, revised at your direction, and published by you.
• Drafted by the assistant and published automatically without a human reading it first — auto mode says exactly that, in those terms.

This page is the Platform’s standing disclosure that the feature exists. The per-post line is the disclosure that matters to a reader, and it is not optional.

You Are Responsible for What You Publish

Plain English: pressing publish makes it yours, however it was drafted — and switching on auto mode is pressing publish in advance. Content published on your blog is your content under the Platform Terms, with everything that means: accuracy, legality, our content rules, and the income-claims policy all apply to it, whether you typed every word, edited a draft, or let auto mode post it while you slept. “The AI wrote it” is not a defense to anything. Verify anything factual before it goes out, because AI drafts can be confidently wrong — and in auto mode, nobody is checking but you, afterwards.

Quality Gates

Drafts are checked before they ever reach you: an objective quality gate on length, structure and machine-tell language, and an originality check against what has already been published on your blog and on other member blogs in your categories. A draft that fails is discarded and regenerated, and a discarded draft does not cost you one of your monthly drafts.

Copyright in AI Output

Plain English: the law may not protect raw machine output, so make the work your own. We claim no ownership of drafts the assistant produces for you; as between you and us, they are yours to use. But under current United States law, content generated by AI without meaningful human authorship may not qualify for copyright protection at all. If someone copies a post you published as raw AI output, you may have no copyright to enforce. Substantial human editing, selection, and addition is what creates protectable authorship. This is one more reason the honest practice, editing drafts into your own work, is also the self-interested one — and it is a real cost of running auto mode.

No Earnings Claims

Plain English: the assistant will not write claims about money results, period. The assistant is built to refuse requests to draft earnings claims, income projections, or “results” copy about money, whether about you, our products, or anything else. This is deliberate and it will not be relaxed. Our Income Disclaimer and the income-claims policy in the Platform Terms explain the standard that applies if you write such claims yourself.

How Your Words Reach the AI

Plain English: your brief goes to an outside AI provider to be processed; other members’ posts do not. When a draft is generated we send your brief, recent post titles from your own blog, and where it is relevant, passages from David Wood’s own recorded training library, to our AI service provider under contractual confidentiality terms. Those providers are currently Anthropic and OpenAI, and which one runs a given draft depends on the model configured at the time; that list can change as the market does and this page will change with it.

The originality check compares your draft against published posts from your blog and from other member blogs in the same categories. That comparison runs on our own servers, mathematically. Those posts are not sent to the AI provider. We do not use your content to train foundation AI models, and we do not permit our providers to.

Labelling Beyond This Page

The per-post disclosure described above is applied by the Platform and you do not have to do anything to get it. Whether you want to say more about how a particular post was made is your call and, in some jurisdictions and contexts, may be your legal obligation. When in doubt, disclose.

Last updated: August 2026.