I'm not putting this in a FAQ line, because it isn't a footnote. Romintel uses AI — Claude, specifically — to do the reading I physically can't: scanning the top 50 bestsellers across eight romance sub-genres every morning, 400 books a day, and tagging what trope, mood and cover style each one carries. It also helps assemble the raw structure of a report you buy — your comps, your categories, the shape of the data.
That's what it does. Here's what it doesn't: it doesn't decide which trend actually matters this week. It doesn't write a word of the commentary in the Briefing, or a word of the copy on this site. It doesn't design anything — no covers, no brand, no reference imagery. Every judgement call is mine. I read the data myself, decide what's worth flagging, and write down what I think it means.
If a verdict on this site sounds like it has an opinion, that's because it does — mine, not a model's.
Most romance readers say they'll happily read an AI-assisted book if it's good. But a smaller, organised part of this community keeps track — by author, by offence — and a cover that merely looks AI-generated gets treated as proof of the worst assumption about everything else in the book, whether or not that assumption holds up.
Romintel visibly uses AI. Staying quiet about that doesn't read as neutral in this environment — it reads like hoping nobody asks. So I'm answering the question before anyone has to ask it.
The words on this site, the Briefing every week, the brand and its look — all written and built by me, by hand, in Johannesburg. AI reads the market so I don't have to read 400 books a day myself. It doesn't write, design, or decide anything you see here.
If that ever changes, I'll say so here, on this page, first.