Science journalist Linda Geddes impersonated in AI-assisted pitch scheme

British science journalist Linda Geddes has described how an impostor used her name to pitch stories to major publications and placed an article under her byline in the New York Post. In a first-person account published on 17 October 2026, Geddes said many of the fake pitches showed signs of having been written with AI assistance.

The scheme surfaced when a friend working as an editor at Nature asked Geddes whether she had been pitching to colleagues. She had not. Someone using an email address that did not belong to her had approached editors at Nature, Science, Nature Medicine and Nature Health, citing her earlier work to appear credible. The Science pitch was rejected.

Geddes then learned of the New York Post piece when a LinkedIn user thanked her for including them in an article on "chronoworking", the practice of organising work around one's body clock. She had never spoken to that person. According to Geddes, apart from a single email interview, the quotes had been taken without credit from previously published features. A New York Post spokesperson said the article was removed, reviewed internally and referred to the relevant authorities.

A possible second impostor

Separately, someone posing as Caroline Williams, a former colleague of Geddes, persuaded a Science editor to commission and publish an article in her name, earning about $1,000. When Geddes ran her impersonator's address through Outlook.com's account recovery tool, it revealed a partly hidden Gmail recovery address; a Science editor had found a matching pattern tied to the fake Williams account, hinting at a link between the two cases.

Simon Miller, director of policy, strategy and communications at UK fraud prevention service Cifas, called the case unusually small-scale for fraud, which typically relies on industrialised methods. Still, he warned: "AI is allowing criminals to create sophisticated impersonations at unprecedented speed and scale."

A wider pattern of stolen names

Prof Eric Topol of the Scripps Institute said that in April 2026 he found his name on a scientific paper he knew nothing about, and that at least 12 books he did not write are sold on Amazon under his name. Leslie McIntosh of Digital Science said respected names are sometimes added to papers or bogus journals without permission to borrow credibility and attract publication fees.

Jamie Collier of Google Threat Intelligence said criminals have used stolen identities to land remote jobs, with large language models producing convincing CVs and fabricated experience, and attackers already testing live deepfakes to pass video screenings. The Society of Authors and the Writers' Guild of Great Britain have also warned of increasingly sophisticated scams impersonating literary agents and publishers.