Effective 2026-04-21 · Version 1.0

The Charter

The Responsibility Ledger exists on one premise: in an era of increasingly powerful AI systems and increasingly confident commentary about them, readers have no reliable way to distinguish signal from marketing. This charter is our answer to that problem.

What this ledger will do

1. Publish daily, dated entries. Each entry carries the date it was written in UTC. No entry is ever backdated.
2. Append-only. Once an entry is published, it is not deleted. Corrections are added as new, dated notes appended to the original entry. A visible Correction block explains what changed and why.
3. Name every source. Every factual claim links to a primary source or names the model, prompt, and output that produced it. If we used an AI system to draft or analyze, we say so.
4. Score prior projections against reality. Every forward-looking claim is logged with a confidence level and an explicit invalidator. When time passes, we grade ourselves in public.
5. Publish a refusal line. Each entry states what the editor refused to do to get you to read that entry.

What this ledger will not do

1. No pay-to-say. We will never accept payment to include, exclude, soften, or elevate a claim. Sponsorships — if any appear — will be clearly labeled and will never influence editorial content.
2. No silent deletions. We will not quietly remove entries, even entries we later find embarrassing. Mistakes are part of the ledger.
3. No fabricated sources. If we cannot cite it, we flag it as unsourced. We will not invent a URL to sound more credible.
4. No anonymous AI output passed off as human analysis. When an AI system produced the text, we say which system, when, and with what prompt.
5. No hidden trackers. This is a commercial publication and we use analytics, conversion tracking, and session replay. What we won't do is hide them. Every tool we run is listed on our Privacy page by name, with what it collects and how to opt out. When we add or remove a tool, we date-stamp the change. Trackers are disclosed, not denied.

The refusal line

Every entry ends with a refusal line — a single sentence describing what the editor chose not to do while writing that entry. Examples:

• Refusal: I did not use a headline that implied certainty I did not hold.
• Refusal: I did not cite a source I had not personally opened.
• Refusal: I did not trim a quote in a way that changed its meaning.

The refusal line is not decoration. It is a discipline: if we cannot honestly write one for an entry, that entry does not ship.

How corrections work

If an entry contains an error, a Correction block is appended to the bottom of the entry with the date, the original sentence, the corrected sentence, and a short explanation. The original text remains in place, struck through. Readers can compare both versions. No URL changes; the same link that returned the error now returns the error plus the correction.

How this is monetized

The ledger is free to read. Monetization, if and when it arrives, will follow these rules:

• A paid tier may offer deeper weekly or monthly analysis. The daily entry always remains free.
• An enterprise license may offer organizations the right to publish their own internal Responsibility Ledgers using our methodology. What we sell is methodology, not endorsement.
• Sponsorships, if accepted, appear in a fixed, clearly-labeled slot and never on editorial pages. A sponsor cannot be the subject of an entry they sponsor.

Who writes this

The Responsibility Ledger is published by Number One Son Software Development, a solopreneur studio. The named editor on each entry is accountable for every claim in it. When an AI system contributed, it is named alongside its prompt and its role.

Versioning this charter

This charter is itself append-only. Changes create a new version, linked from every prior version. The current version and its effective date are shown at the top of this page.

Last updated: 2026-04-21 · Version 1.0 · Signed: Roger Grubb, Editor