APS has revised its AI policy to expand permissible AI use in manuscript preparation and peer review, while maintaining a hard line on human oversight and accountability. This is a meaningful signal: a major scientific society formally accommodating AI as a legitimate authorship/review aid rather than treating it as a blanket threat. The 'expand permitted use + reinforce human oversight' framing is becoming standard language — worth tracking how it plays out in practice versus policy. Compare against Frontiers' AI guidance (April 2026) and the STM responsible use framework.

Source: Knowledge Speak