Policy on AI and AI-Assisted Technologies

Primary Care Science and Practice (PCSP) recognizes the potential of generative AI in scholarly writing. However, to maintain the highest standards of transparency and integrity, we adopt the following policies for Authors and Reviewers:

1. Guidelines for Authors

Authorship Eligibility: Generative AI (such as LLMs, chatbots, or image creators) cannot be listed as an author or co-author. Authorship implies legal and ethical accountability that AI cannot fulfill.

Permitted Use: AI may be used to improve the readability, grammar, and language accuracy of the manuscript. However, all scientific content, logic, and conclusions must be the work of the human authors.

Generative AI Images: The use of generative AI to create or alter scientific images, figures, or charts is prohibited. Any exception (e.g., if AI is the subject of the research) must be justified and described in the Methodology section. Original data visualization must remain untouched by generative filters.

Mandatory Disclosure: Authors must disclose any use of AI in a dedicated "Declaration of AI Use" statement before the References or within the Methodology/Acknowledgments section.

Required Details: Name of the model (e.g., GPT-4o), version, date of use, and a specific description of how it assisted in the work.

Responsibility: Authors are solely responsible for the accuracy of their work. AI-generated text that contains misinformation or plagiarism is the sole liability of the authors. AI should not be cited as a primary source.

2. Guidelines for Reviewers

To protect the confidentiality of the authors' intellectual property, PCSP implements a strict policy for Peer Reviewers:

Confidentiality: Reviewers are prohibited from uploading any part of a submitted manuscript into generative AI tools (such as ChatGPT or Gemini). Doing so constitutes a breach of confidentiality and potentially compromises the author's original data.

Review Integrity: AI-generated peer review reports are discouraged. While AI may assist in checking specific data or grammar, the evaluation of scientific merit, clinical relevance, and ethical compliance must be conducted by the human reviewer.

Disclosure: If a reviewer uses AI to assist in the review process (e.g., for data verification), this must be disclosed to the editor.