Review of Finance bans hidden prompts in manuscripts and sets a one-year penalty

The Review of Finance, the journal of the European Finance Association, published a policy on large language models on 23 March 2026 and updated it on 23 June 2026. It covers referees and authors.

For referees, the policy says LLMs may be used in a supportive way but must not substitute for the referee’s own evaluation, interpretation or recommendations.

For authors, LLMs must not replace intellectual contribution, analysis or interpretation. AI tools may be used in a supporting role, for example to improve language, assist with coding or support data handling, and that use must be transparent. The journal follows Oxford University Press ethics guidelines: if AI tools were used in any capacity, this must be disclosed in the cover letter and acknowledged in the manuscript, for example in the Methods or Acknowledgements section.

The policy also prohibits submitting any manuscript file that contains hidden text, altered character encodings or embedded software instructions designed to manipulate automated systems. A first violation means rejection without refund and a one-year suspension of the authors’ submission privileges. Later violations are sanctioned at the Managing Editor’s discretion.