The courts have acknowledged that medical documentation need not meet a standard of perfection. In Brito et al. v. Woolley et al., the court held that “the law does not impose a standard of perfection on medical personnel in their preparation and maintenance of medical records,” recognizing that “[o]ccasional inconsistencies, inaccuracies, and/or omissions are tolerated.”20 This precedent assumes, however, that the records were generated by the treating physician or someone under their direct supervision, not by a third-party algorithm that may incorporate biases or import false information.
This assumption becomes even more critical when considering evidentiary rules. Section 42 of the British Columbia Evidence Act governs the admissibility of medical records as business records in legal proceedings.21 Among other requirements, this section stipulates that the weight to be given to such documentary evidence is affected by whether it was created by a person with personal knowledge of the matters being recorded. If AI is used to generate content that the physician has not verified or that does not accurately reflect a direct recollection or observation, the reliability and the weight to be accorded to such documentation may be challenged.
It follows, then, that medical malpractice lawyers must be vigilant when reviewing medical records that may have been generated by digital scribes. It is essential to assess whether the documentation reflects the physician’s direct observations and clinical reasoning, or whether it has been shaped by algorithmic structuring that misrepresents the actual encounter. Lawyers should inquire whether the record was created using AI, whether informed patient consent was obtained for recording and transcription, and whether the physician thoroughly reviewed and verified the AI-generated content before signing off. In cases where the record appears central to the alleged negligence, counsel should consider whether the document meets the evidentiary requirements under section 42 of the Evidence Act.
Conclusion
As the use of AI in charting becomes more common, scrutiny of how these tools are deployed and documented will become an increasingly important aspect of effective malpractice advocacy. Medical malpractice lawyers must be alive to this fact and should actively investigate whether AI-generated notes were properly reviewed and supervised by the treating physician, whether clinical decisions were unduly influenced by algorithmic outputs, and whether the documentation faithfully reflects both the patient’s condition and the physician’s own clinical reasoning.
Incursions by artificial intelligence are no longer a distant or hypothetical concern for the medical malpractice lawyer; they are a present and pressing reality. Understanding the role that AI may have played in the generation of a medical record is now essential not only to assessing liability, but also to safeguarding the integrity of the record itself as a piece of legal evidence.
