ChatGPT Anthropomorphism Spurs Publishing Authorship Concerns

The London Review of Books blog post "Chattiness" documents cultural shifts in how readers and writers engage with ChatGPT, noting frequent anthropomorphizing and routine use for tasks from restaurant recommendations to drafting reports. The post reports that Hachette bought the rights to *Shy Girl* by Mia Ballard, released it in November, and later pulled it from publication in the UK and US, after a Reddit thread by a self-identified "book editor of twelve years" flagged passages that readers perceived as AI-like (per the London Review of Books). The post says Ballard denied using AI and blamed a freelance editor (per the London Review of Books). The blog also recounts online doubts about the Commonwealth Short Story Prize Caribbean winner by Jamir Nazir and notes that the Commonwealth Prize told the piece that entrants had affirmed their work was their own (per the London Review of Books).

Industry-pattern observations: publishers and readers are encountering generative text in creative workflows more frequently, creating friction around provenance, editorial credit, and detection. Current automated-detection tools trade off false positives and false negatives; independent reporting and community flagging remain a common first signal. For practitioners, this increases demand for robust provenance metadata, watermarking experiments, and reproducible editorial audit trails rather than ad hoc inspection alone.

Observed patterns in similar episodes show that public controversies often combine stylistic markers that communities recognize with gaps in publishing vetting processes. This combination elevates reputational risk for publishers and complicates standard editorial workflows. Publishers and platforms are therefore under pressure from readers and peers to produce clearer submission provenance and to document editorial interventions.

Indicators an observer can follow include:

• •whether major publishers adopt mandatory provenance or disclosure policies for submitted manuscripts;
• •adoption of cryptographic watermarking or metadata standards by commercial model providers and publishers;
• •regulatory or industry guidelines addressing attribution and liability in published AI-assisted creative work.