AI Use Policy

This page documents how The Ledgerism Brief uses artificial intelligence in editorial production, what we never use AI for, and how we disclose AI involvement when it occurs.

What we use AI for

We use AI tools to assist with:

• Initial research and source location (locating primary documents, finding court records, identifying which regulatory body issued which release)
• First-pass copy editing and voice gate enforcement (catching em-dashes, banned phrases, passive constructions)
• Image generation for editorial illustrations and data visualizations (Replicate Flux 1.1 Pro)
• Headline variation generation during the editing process
• Code generation for tooling that supports editorial production (this site is built with AI-assisted code)

What we never use AI for

We never use AI to:

• Write the substantive content of an article from a prompt without human authorship
• Generate quotations or attributions
• Generate citations to court cases, regulations, or accounting standards (every citation is human-verified against the primary source)
• Generate research data we then publish as our own findings
• Generate firm rankings, partner names, or credential claims
• Respond to reader emails or correction requests

Disclosure

Every article on The Ledgerism Brief is written by a named human author, reviewed by a named human editor or peer reviewer, and the byline is real. If we ever publish AI-generated content in any substantive form (which is not our current practice), it will be clearly labeled at the top of the page in a way that cannot be missed.

Why we publish this policy

The accounting profession has a higher trust standard than most other professions. A CPA’s signature on an audit report has a defined legal meaning. We hold ourselves to a corresponding standard on what we publish. Disclosing where AI helps and where AI is never permitted is the minimum we owe readers who rely on our reporting.

Policy review

This policy is reviewed every 90 days. Material changes will be noted with a dated update at the bottom of this page.

Policy version 1.0, effective June 2026.