AI scribes are getting drug names and diagnoses wrong in NHS records

A patient in England was told they had demyelination, the nerve damage that underlies conditions including multiple sclerosis. The test result had actually read “null demyelination”, and an AI scribe had dropped the word that reversed the meaning.

That case appears in a warning from Healthwatch England, the statutory patient watchdog, reported by the Guardian today. Its finding is that the tools now transcribing consultations for GPs and hospital doctors are getting drug names and diagnoses wrong, and that patients are often the ones who notice.

Twenty-seven different AI scribes are in use across the health service in England. They sit in the consulting room, listen, and produce the note that goes into the record and the letter that goes to the patient.

The errors collected by the watchdog are mundane in form and serious in effect. One scribe swapped a prescribed drug for a different one with a similar name, the kind of confusion that pharmacology spends considerable effort trying to design out.

Another summary omitted a consultant’s instruction that the patient should seek a repeat prescription for migraine medication. A third recorded a doctor telling a patient to continue their Prozac, when that doctor had neither prescribed it nor discussed it.

The common thread is that nothing looked broken. A fluent, plausible note is exactly what these systems produce, which is why an error survives the glance a busy clinician gives it before signing off.

“These inaccuracies may persist in their records if the patient doesn’t catch them,” the watchdog warned. That places the last line of defence on the person least equipped to know what the note should have said.

Rachel Power, chief executive of the Patients Association, is among those raising concerns, alongside clinicians including the London GP Shier Ziser Dawood and Charlotte Blease of Uppsala University in Sweden. The objection is not to the technology so much as to its arrival without a safety net.

Because there is no England-wide oversight of these tools. The Medicines and Healthcare products Regulatory Agency has not classified AI scribes as medical devices, which leaves them outside the regime that would test them for safety and effectiveness before deployment.

The regulator did publish guidance in August clarifying where the line sits. A system that only transcribes what was said is not a device, while one that suggests a diagnosis or a treatment may well be, which puts a great deal of weight on how each product is described by its vendor.

The incentive that follows is obvious enough. A scribe marketed as a passive transcriber avoids a regulatory process that a scribe marketed as a clinical assistant would have to complete.

Transcription error is also a different failure from the one most AI safety work anticipates. Nobody here was misled by a hallucinated fact; a real sentence was rendered slightly wrong, and slightly wrong is sufficient when the sentence names a drug.

None of which addresses why these tools spread so quickly. Clinical documentation is the administrative burden doctors complain about most, and a system that reliably removes an hour of typing a day will be adopted whether or not anyone has assessed it.

The problem is what happens in between those two facts. A tool adopted for its speed, unassessed because of how it is categorised, producing a document that becomes the permanent clinical record, is a chain in which no single link is obviously anyone’s responsibility.

The remedy the watchdog points to is unglamorous and probably right. Patients should be told when a scribe is being used and given their notes to check, which turns an accidental safety mechanism into a deliberate one.

Twenty-seven products, no device classification, and a check performed by whoever happens to read their letter carefully: that is the current arrangement in the NHS in England.