Re: Six rules for doctors using AI

Agreement: I Agree Body: Dear Editor Daniel Sokol⁽¹⁾ provides practically valuable guidance for clinicians on the real-world use of artificial intelligence in clinical practice. He clarifies the dual pitfalls of over-reliance on and outright rejection of AI, and defines physicians’ primary responsibilities and operational principles. Nevertheless, the article places excessive emphasis on individual clinician accountability while downplaying institutional and systemic-level challenges. Current risks associated with clinical AI stem not only from inappropriate physician performance but also from deeper structural issues including lagging industry regulation, absent algorithmic standards, and ambiguous technical-access criteria⁽²,³⁾. By disproportionately assigning fallback liability to frontline doctors, the paper overlooks the shared obligations of developers, healthcare organisations and regulatory authorities. Furthermore, it does not address systemic clinical harms arising from algorithmic bias and skewed training-data biases — risks that cannot be mitigated solely through clinicians’ due diligence⁽⁴,⁵⁾. Accordingly, regulating clinical AI implementation should go beyond constraining individual clinical behaviour. Establishing harmonised frameworks for AI medical product access, algorithm auditing and liability attribution, alongside improved algorithm governance and professional standards within a multi-stakeholder collaborative model, is essential to mitigate clinical hazards at their source and enable safe, beneficial technological empowerment of healthcare. References 1.Sokol D. Six rules for doctors using AI. BMJ. 2026;394:e100883. doi:10.1136/bmj 2026 100883. 2.Warraich HJ, Tazbaz T, Califf RM. FDA perspective on the regulation of artificial intelligence in health care and biomedicine. JAMA. 2025;333(3):241 247. doi:10.1001/jama.2024.21451. 3.Lekadir K, Frangi AF, Porras AR, et al; FUTURE AI Consortium. FUTURE AI: international consensus guideline for trustworthy and deployable artificial intelligence in healthcare. BMJ. 2025;388:e081554. doi:10.1136/bmj 2024 081554. 4.Abdulnour Raja Elie E, Gin B, Boscardin CK. Educational strategies for clinical supervision of artificial intelligence use. N Engl J Med. 2025;393(8):786 797. doi:10.1056/NEJMra2503232. 5.Mahase E. AI: doctors risk being sued if tools go wrong, while companies are “shielded,” report warns. BMJ. 2026;393:273474. doi:10.1136/bmj 2026 273474. No competing Interests: Yes The following competing Interests: Electronic Publication Date: Wednesday, October 7, 2026 - 08:56 AI use: No, I have not used AI Highwire Comment Subject: Six rules for doctors using AI Workflow State: Released Full Title: Re: Six rules for doctors using AI Highwire Comment Response to: Six rules for doctors using AI Check this box if you would like your letter to appear anonymously:: Last Name: Shi First name and middle initial: Jizhou Email: jizhou_shi@163.com Address: No.31 Jinan Road, Dongying District, Dongying 257034, Shandong Province, China Occupation: Consultant surgeon Other Authors: Mingxin Wang, Consultant surgeon, Shengli Oilfield Central Hospital Affiliated to Shandong Medical and Pharmaceutical University Affiliation: Shengli Oilfield Central Hospital Affiliated to Shandong Medical and Pharmaceutical University BMJ: Additional Article Info: Rapid response