This is a delicate situation. You need to be professional, factual, and restrained. Do not openly accuse the reviewer of “using AI” unless you have irrefutable proof (which is rare). Accusing a reviewer of using AI can backfire if you are wrong or if it comes off as an attack.
Instead, focus on the symptoms of AI use: hallucinations, factual errors, generic/vague text, and lack of engagement.
Here is a draft email to the Area Chair (AC).
Subject: Inquiry regarding [Paper ID XXXX] discussion and review quality
Dear Area Chair,
I hope this message finds you well.
I am writing to respectfully bring two matters regarding the review process of our submission [Paper ID] to your attention as the discussion period draws to a close.
1. Lack of Reviewer Engagement
We have provided detailed responses, new experiments, and theoretical corrections to address the concerns raised by the reviewers. We also posted a polite follow-up reminder recently. Unfortunately, we have not yet received any replies or acknowledgement of these revisions from Reviewers [list unresponsive Reviewer IDs, e.g., GuXF, SGM9]. We are concerned that our substantial updates (including new baselines and quantitative XAI metrics) may not be factored into the final decision if the reviewers do not re-engage.
2. Concerns regarding Reviewer [ID of the ‘AI’ reviewer]
We also wish to flag a concern regarding the review provided by Reviewer [ID]. We noticed several factual inaccuracies in their critique that do not align with the content of our paper or standard literature. For example:
- [Specific Hallucination 1]: The reviewer claimed [X], but our paper [does/says Y].
- [Specific Hallucination 2]: They cited a “standard” that does not exist/is not applicable.
The feedback was notably generic and contained statements that appear to be hallucinations. We have addressed these points factually in our rebuttal, but given the nature of the errors, we wanted to ensure the AC is aware of potential quality issues with this specific report when weighing the final scores.
We understand that this is a busy period for everyone, but we would greatly appreciate your oversight to ensure a fair evaluation of our revised manuscript.
Thank you for your time and service to the conference.
Sincerely,
The Authors of Paper [ID]
💡 Why this framing works:
- “Factual Inaccuracies” vs. “Used AI”: ACs know that “hallucinations” + “generic text” = AI. You don’t need to say the words “ChatGPT.” Letting the AC connect the dots is more professional than making an accusation.
- Focus on Fairness: You are framing this as a request for a fair evaluation of your revisions, not just complaining that people are ignoring you.
- Actionable for the AC: This prompts the AC to look specifically at that one reviewer’s report with skepticism, which is exactly what you want.