Some sections (Tori and Mike in the living room) are pretty obviously chatGPT/LLM generated, and not in a good way. They're interspersed with what is probably human writing, but goddamn do they kill the flow. Given the pretty consistent grammar/spelling errors throughout the human authored sections (and the very strange idea of what the CIA and navy seals actually do lol), I'm guessing that english isn't the author's first language (and they're certainly not American, the storytelling feels very Russian if anything), so my honest opinion is that LLMs were used because maybe that style of writing is less obviously awful to non-native speaker (or maybe it even seems 'smart' since it uses tons of purple prose) and not because the author was trying to sneak anything by or was just lazy.
The human writing is good, please stop with this junk. At the very least, edit it so it actually makes sense (and use a smarter model).