Commentary Research and Evidence Ethics, Regulation, and Responsible Use

Stop pretending AI peer review isn’t happening

April 21, 2026 By Editor-in-Chief: Johnson Thomas, MD, FACE, FEAA 8 min read
Share Share via Email Share on Facebook Share on LinkedIn Share on Twitter

Clinical Scorecard: Stop Pretending AI Peer Review Isn’t Happening

At a Glance

CategoryDetail
ConditionPeer Review System Challenges
Key MechanismsIntegration of AI to enhance peer review quality and efficiency.
Target PopulationResearchers and academic journals facing peer review bottlenecks.
Care SettingAcademic publishing and scientific evaluation.

Key Highlights

  • AI can improve peer review quality without biasing outcomes.
  • Human reviewers are overwhelmed, with 10% handling half of all reviews.
  • AI excels at objective assessments, freeing human reviewers for subjective judgments.
  • The ICLR 2025 experiment showed AI feedback led to improved review quality.
  • Current policies largely prohibit AI use in peer review despite its potential benefits.

Guideline-Based Recommendations

Diagnosis

  • Identify inefficiencies and biases in the current peer review process.

Management

  • Implement AI tools for pre-screening and feedback in peer reviews.

Monitoring & Follow-up

  • Evaluate the impact of AI on review quality and acceptance rates.

Risks

  • Concerns about confidentiality and AI 'hallucinations' must be addressed.

Patient & Prescribing Data

Scholars and researchers involved in the peer review process.

AI can provide structured feedback to enhance review quality.

Clinical Best Practices

  • Adopt a collaborative model combining AI and human reviewers.
  • Utilize AI for systematic checks while reserving subjective evaluations for humans.
  • Ensure secure implementation of AI tools to protect confidentiality.

References

AACE Endocrine AI is published by Conexiant under a license arrangement with the American Association of Clinical Endocrinology, Inc. (AACE®). The ideas and opinions expressed in AACE Endocrine AI do not necessarily reflect those of Conexiant or AACE. For more information, see Policies.

Related Content