# Post ideas - 2026-10-12 Based on AI-and-writing news from roughly Oct 5-12, 2026. ## 1. The Ghostwriter in the Publisher's Basement **Angle:** Wired reports that three of the Big Five are quietly using AI for blurbs, cover art and publicity copy, so memoirists should ask where the line sits between "tool" and "author's voice" in their own contracts. - Anonymous sources (unverified) say publishers use ChatGPT/Claude for cover copy and image generators for cover art, without telling authors. - Agents are reportedly now demanding contract clauses that bar publishers from feeding unpublished manuscripts into LLMs. - Memoir angle: your jacket copy and your life story are both "your voice". Who gets to paraphrase it? - A practical checklist: clauses to ask for, and questions to put to a publisher or small press. **Sources:** - [Gizmodo, Oct 10 2026 (relaying Wired)](https://gizmodo.com/as-they-supposedly-battle-ai-slop-major-publishers-reportedly-cover-their-books-in-it-2000824603) ## 2. "Prove You Wrote It": The $2 Million Memoir Problem **Angle:** A reported $2M+ novel deal was dropped over doubts about AI authorship, so writers, memoirists especially, may soon need a paper trail proving the work is theirs. - Minotaur Books reportedly pulled a roughly $2-2.4M crime-novel deal over suspected AI content; the article notes the concern was less about quality than about copyright in AI-generated elements. Coverage was largely from July/Aug and resurfaced in this week's reporting. - Accusation without proof is the new risk: detection tools are unreliable and the debate is already heated (see Substack's AI-detection backlash). - Build a "provenance habit": dated drafts, notes, voice memos, version history in Obsidian. - Turn it into a web project: a simple "how I wrote this" page or badge for your own work. **Sources:** - [Gizmodo, Oct 10 2026](https://gizmodo.com/as-they-supposedly-battle-ai-slop-major-publishers-reportedly-cover-their-books-in-it-2000824603) - [Plagiarism Today, Aug 5 2026](https://www.plagiarismtoday.com/2026/08/05/author-loses-2-million-book-deal-over-ai-allegations/) - [404 Media on Substack's AI detector "witch hunt"](https://www.404media.co/substackers-say-new-ai-detection-tool-is-a-witch-hunt/) ## 3. Check Whether Your Book Is in the Pirate Library (and What the Lawsuits Mean for Your Archive) **Angle:** With the Anthropic settlement in the news again and USA Today Co. suing OpenAI for $250M+ over training data, working writers should audit where their own words are exposed and decide what they want to license or block. - Bartz v. Anthropic: a roughly $1.5B settlement for pirated books; the court held training on legally acquired books was fair use, but a pirated "central library" was not. Settled, so no precedent set. - On Oct 8, USA Today Co. and its newspapers sued OpenAI for $250M+, alleging hundreds of thousands of articles were used for training. - Takeaway for a writer with web projects: check the settlement lists for your titles, review your robots.txt/AI-crawler policy, and decide on a licensing stance for your site. - Provocative question: is a blog archive worth more as training data or as a moat? **Sources:** - [Tufts Daily, Oct 8 2026](https://tuftsdaily.com/article/2026/10/tufts-professors-book-among-illegally-pirated-works-in-anthropic-copyright-settlement) - [Crypto Briefing on USA Today Co. v. OpenAI, Oct 8 2026](https://cryptobriefing.com/usa-today-sues-openai-copyright-250m) - [NPR on the Anthropic case](https://www.npr.org/2026/07/27/nx-s1-5904606/anthropic-vs-bartz-ai-copyright-lawsuit-pros-cons)