A working paper from the National Bureau of Economic Research found that nearly 30% of STEM PhD dissertation abstracts submitted in the US this year contained AI-generated text, up from under 2% in 2023. The study, which analyzed over 226,000 dissertations using the detection tool Pangram, also found that graduates whose dissertations showed AI-generated text were more likely to go into industry jobs and less likely to pursue academic careers.
nature.com
· 2026-10-16
A new essay argues that writers should draft their own work entirely by hand, then feed it to a large language model purely to spot flaws, never to supply finished phrasing. The author lays out two guiding rules: never use a specific wording an LLM suggests, and treat the model strictly as a copyeditor rather than a co-writer, to avoid producing text that reads as generic machine output.
sockpuppet.org
· 2026-09-17
Tools such as Pangram and Originality.ai are being used to identify AI-generated writing, offering some clarity for educators, publishers and employers trying to spot machine-authored content. However, these detectors remain imperfect, producing errors that limit how much confidence users should place in their results.
gizmodo.com
· 2026-09-09