AI policy.

The short version

The website you are looking at today was primarily built with the aid and assistance from Claude Code (an AI programming assistant). This includes the page structure, the SRS algorithm installation, and getting everything formulated to be a proper running site, along with more details listed below. That said, the curriculum and all text on this site was not. Every module, every citation, every claim about how American healthcare works, and certainly every terrible joke along the way was researched and written by me (Jackson White). I've calculated my AI usage for this site to be roughly ~8 gallons of water maximum at the time of building (check out this calculator here). While it's not much, I've gone ahead and pledged $100 to a charity, but I do need some ideas on the best ones to donate. I'll finalize the decision by November 1st, 2026 and research good options before then, but please feel free to reach out to me at jackson[at]wtlgovernance.com with some ideas if you'd like to share.

Where AI was used

As mentioned previously, Claude Code was used throughout the engineering of this site. This includes application code, the content pipeline, the quiz and spaced-repetition systems, the database schema, styling, deployment configuration and the drafting of the privacy policy and terms pages. Where it produced something, I personally reviewed it before it shipped.

Where AI is not used

I did not (and will not) use AI to write the curriculum. Specifically, it is not used to draft module explanations, to summarize statutes or regulations, to generate citations, to produce quiz questions, or to decide what a rule means. For better or worse, those come from me directly.

The reason is three-fold. First, this site was a building block for my own education. By going through the research process and building this site, I've had the opportunity to educate myself in a way that would've been impossible had I just asked AI to do it for me. Second, my goal with the site is to provide a unique, engaging, and fundamentally human learning experience. If everything was put in the style of "AI slop" writing, it would fail to be engaging, which would defeat the entire purpose. Third, and perhaps most important, I am insufferable and love the sound of my own voice (or writing in this case).

How sources are handled

As I structure the course, modules are written by pulling in public primary sources (e.g. statutes, regulations, agency manuals, public datasets, and published enforcement actions). Each module lists the authorities it relies on and links to them, so you can check the claim rather than take it on my word. Each also carries the date it was last verified against those sources. With the ability to comment on each module, please yell at me if I got something wrong, whether in the sourcing, the interpretation, or if the information is out of date. I promise not to cry (too much).

If this changes

Should AI ever be used in preparing curriculum material (for example to draft practice questions from text a human has already written, or to flag modules affected by a new rule), you can have my legal and express opinion to cyberbully me.

Your contributions

I've set this up to be a manual approval system so that no spam going to shady sites starts popping up here. But for the most part, I will flag comments to go public, especially if they're helpful to the community at large.

Corrections

If you find an error, please call it out. I've got the spelling of a rabid kindergartner, and while auto-correct sometimes helps, it isn't foolproof. More importantly, if you notice any content, links, or interpretations are incorrect or outdated, please post a comment. Corrections are recorded in that module's revision history.