Academic Publishers Clash Over AI's Role in Research at Bentley University Panel

Academic publishers and journal editors clash at a Bentley University panel over Generative AI's role in research, raising urgent concerns about hallucinations, bias, intellectual property risks, and 'scholarly deskilling,' while agreeing on disclosure but failing to reach unified standards.

• A panel held in October 2025 at Bentley University brings together major publishers and journal editors to debate the growing use of Generative AI in academic research, revealing deeply divided opinions on how far AI should be embedded in the research process.
• Key concerns raised include AI hallucinations producing inaccurate outputs, inherent dataset biases, intellectual property risks during peer review, and the threat of 'scholarly deskilling' as researchers increasingly delegate cognitive tasks to automated systems.
• Publishers and editors agree that full disclosure of AI use is urgent and that outsourcing peer review to AI is inappropriate, but inconsistent policies across institutions and the lack of reliable AI detection tools leave the academic community without clear, unified standards.