How this site uses AI

This tracker is about lawyers using AI responsibly, so here is our own disclosure. AI research agents help find candidate instruments and cases, draft plain-English summaries, and build the site itself. No legal claim ships because a model produced it: every instrument, date, and citation is verified against its primary source, the issuing bar, court, or the ABA, before it appears here, and an automated verification gate blocks any release that breaks those rules.

What AI does here

Research agents built on large language models carry much of this site's production: sweeping for newly issued opinions, rules, guidance, and sanctions decisions; drafting summaries and state pages; writing the site's code; and keeping the tables, datasets, and endpoints consistent with one another. The AI-readable index at /llms.txt and the full plain-text dataset at /llms-full.txt are generated from the same verified data as the human-readable pages on every build, so they cannot drift from what the site actually says.

What AI is never allowed to do

No AI output becomes a legal claim on this site without primary-source verification. In practice that means: an instrument is listed only after the primary document has been read; a citation that cannot be confirmed against the primary source is treated as a hallucination and dropped or replaced with a docket citation; dates carry only the precision the primary source supports; and guidance is never upgraded to a "formal opinion" to make a headline count bigger. The methodology page states these rules in full, and the fabrications watch documents what circulates in this space when they are skipped.

The verification pipeline

Behind the site sits a verified fact record and an automated gate that runs before every release. The gate checks, among other things, that the published datasets match the source data field for field, that every date is a real calendar date labeled with honest precision, that citation strings we found to be unverifiable appear nowhere in public output, and that every page carries its disclosures. A human operator reviews and publishes every release; nothing on this site deploys automatically. Every content change lands on the public changelog.

Why we tell you this

The ethics opinions this tracker catalogs converge on one duty above all: verify AI output before relying on it. We hold this site to the same standard and think you should too. MHSB Solutions is not a law firm, and no one on the research desk is a licensed attorney; that is disclosed on every page, and it is why every entry links the primary source so you can check our work. If you find an error, the corrections policy applies: email info@mhsbsolutions.com with a link to the primary source, and verified corrections ship within a week with a changelog entry.

For licensed attorneys and firm operators. This site is legal information, not legal advice, and no attorney-client relationship is formed by using it. Rules change; verify against the primary sources linked on every page and consult a licensed attorney in your jurisdiction before acting.