An illustration of a skeleton as it appears on X-ray

Will AI Read My Scan? How Artificial Intelligence Is Used in UK Radiology, and Who Still Signs the Report

Artificial intelligence is now part of radiology in many NHS trusts, and on 21 October 2026 the Royal College of Radiologists published updated guidance on how AI tools should sit inside the systems radiologists report from. If you are about to have an X-ray, CT, MRI or ultrasound, here is what the software does, what it does not do, and who is responsible for the report you receive.

Written by the FindScan Editorial Team and reviewed by our Medical Advisor, Dr Mohamed Najjar
Reviewed

Ask a radiology department in 2026 whether artificial intelligence is involved in reading scans and the answer in many places is yes, in a limited and supervised way. The Department of Health and Social Care, the Department for Science, Innovation and Technology and NHS England said in June 2026 that AI chest X-ray tools are in about half of England's NHS trusts and will be rolled out to every trust by 2029, and that more than 4 million patients have had a faster lung cancer diagnosis or all-clear with their help.

On 21 October 2026 the Royal College of Radiologists (RCR) published the second edition of its guidance on integrating image analysis AI into the reporting systems radiologists use. So it is a fair moment to ask the question patients ask: will a computer read my scan, and if so, who is responsible for what it says?

What AI in radiology actually does today

Most AI tools in UK radiology are image analysis software that looks at a scan after it has been taken and flags things for the reporting clinician. The RCR describes its new guidance as covering the stage when AI algorithms examine images alongside, and often before, the formal radiology report. The GOV.UK release of 10 June 2026 calls the chest X-ray tools a virtual second pair of eyes for radiologists.

NICE has looked at the two best-known uses, and reached different conclusions. For AI technologies that help detect fractures on X-rays in urgent care (originally HTE20, now HealthTech guidance HTG739), NICE says four technologies can be used in the NHS during an evidence generation period, describes them as used in addition to standard care in which healthcare professionals make treatment decisions, and says they do not replace the definitive radiology review. For AI software that analyses chest X-rays for suspected lung cancer in people referred by their GP (originally HTE12, now HTG696), NICE says more research is needed, and that centres already using it may continue only under an appropriate evaluation framework and only alongside clinician review.

Other uses are growing, such as software that measures structures or helps prioritise which scans are read first. What you should not find in a regulated UK setting is an AI producing and signing a report with no clinician involved.

Who is responsible for your report

The report you or your GP receive is written by a person: a radiologist (a doctor who specialises in interpreting scans) or, for some examinations, a reporting radiographer or sonographer trained for that work. The RCR's guidance is addressed to exactly these people and to the staff who buy and implement AI, and its stated purpose is to help teams add AI to their reporting systems without increasing the reporting burden.

In the GOV.UK announcement, the RCR's president, Dr Stephen Harden, said AI should support doctors by helping them work more effectively, rather than replacing the expert judgement that patients value, and that radiologists will remain at the centre of diagnosis and patient care. In practical terms, an AI flag is a prompt for the clinician to look again, agree or disagree, and the clinician's reading is the one that goes into your record. NICE's fracture guidance makes the same point from the other side: it says clear local protocols are needed so that staff know what to do when they disagree with the software.

The software itself is regulated. The MHRA's guidance on software and AI as a medical device explains that many software products, including AI, are regulated in the UK as medical devices. See radiographer vs radiologist for the human roles in more detail.

Why the NHS is using it

The driver is volume. GOV.UK says more than 7 million chest X-rays are performed across the NHS each year, and the June 2026 release reports early data suggesting faster reporting with the AI tools in place. That is why the first targets have been high-volume, time-pressured tasks such as chest X-rays and urgent care fracture X-rays, where a second check can catch a subtle finding. Evidence on outcomes is still being gathered, which is why NICE's guidance on both uses comes with plans for further evidence. Our guide to fractures an X-ray can miss explains why subtle breaks are a natural target.

What this means for a private scan

Private clinics in the UK operate under the same umbrella: the clinic is registered with the Care Quality Commission in England (or its equivalent in Scotland, Wales and Northern Ireland), X-ray and CT examinations must be justified by a referrer under IR(ME)R 2017, and AI software used as a medical device is regulated. Some private providers use AI tools; many do not.

Either way the report is a human document. If a clinic uses AI, the accurate description is that software assisted the radiologist, not that the scan was read by AI. If you receive a report that mentions AI or a computer-aided detection result, it should be clear which findings the radiologist confirmed. See how to get your scan images and report and how long scan results take.

Questions worth asking any clinic

Who will report my scan, and what are their qualifications? Does the clinic use AI software in reporting, and if so, for which scans? Is the software a UK-regulated medical device? Will the report say whether AI was used? Who do I speak to if I have a question about a finding? Will my images and report be sent to my GP or specialist, and in what format?

None of these questions should be awkward for a good clinic. The RCR's guidance exists because departments are expected to be able to explain how AI fits into their workflow.

The bottom line

AI in UK radiology is real, growing and, so far, assistive. It helps clinicians find and prioritise, it is regulated as a medical device, and a clinician remains responsible for your report. The RCR's updated guidance is about doing that safely, not about handing the job over.

If you need a scan and want to compare clinics, you can browse scan types or find clinics by city and send an enquiry; the clinic contacts you, and you can ask it the questions above at the same time. FindScan is a directory of independent, regulated imaging clinics: you compare clinics and send an enquiry, and the clinic contacts you. We are not a healthcare provider, and nothing here replaces advice from the clinician who referred you. If you have symptoms that need urgent attention, contact your GP or NHS 111, and call 999 in an emergency.

Common questions

Does AI read scans in the NHS?

AI software helps with some scans, mainly chest X-rays and fracture X-rays. GOV.UK said in June 2026 that AI chest X-ray tools are in about half of England's NHS trusts and will reach all of them by 2029. A radiologist or other trained clinician still produces the report.

Can an AI diagnose cancer from an X-ray?

No. NICE's guidance on AI for chest X-rays in people referred by their GP (HTG696, formerly HTE12) says more research is needed and that the software should only be used alongside clinician review. Diagnosis depends on a clinician's reading and, where needed, further scans and a biopsy.

Is AI used for MRI and ultrasound too?

Less often for finding abnormalities. Some software helps with measurement or with how quickly images are acquired, but the uses NICE has assessed for detecting problems are chest X-rays and fracture X-rays.

Will I be told if AI was used on my scan?

Not always. You are entitled to ask, and a clinic should be able to say whether software assisted the report and which findings the radiologist confirmed.

Who is responsible if an AI tool misses something?

The clinician who reports the scan remains responsible for the report. NICE's fracture guidance says the AI tools are used in addition to standard care and do not replace the radiology review, and AI software used as a medical device is regulated by the MHRA.

References

  1. Integrating image analysis artificial intelligence with radiology reporting workflows (RIS and PACS), second edition published (21 October 2026). Royal College of Radiologists. Accessed 27 October 2026.
  2. Integrating image analysis artificial intelligence with radiology reporting workflows (RIS and PACS): second edition (publication page). Royal College of Radiologists. Accessed 27 October 2026.
  3. AI to speed up cancer diagnosis for millions of NHS patients (10 June 2026). GOV.UK (Department of Health and Social Care, DSIT and NHS England). Accessed 27 October 2026.
  4. HTG696 (formerly HTE12): AI-derived software to analyse chest X-rays for suspected lung cancer in primary care referrals: early value assessment. NICE. Accessed 27 October 2026.
  5. HTG739 (formerly HTE20): AI technologies to help detect fractures on X-rays in urgent care: early value assessment (14 January 2025, updated 23 October 2026). NICE. Accessed 27 October 2026.
  6. Software and artificial intelligence (AI) as a medical device. GOV.UK (MHRA). Accessed 27 October 2026.
  7. The Ionising Radiation (Medical Exposure) Regulations 2017. legislation.gov.uk. Accessed 27 October 2026.

This article is general information, not medical advice, and it does not replace an assessment by a clinician. If you have symptoms that worry you, speak to your GP or call 111. In an emergency, call 999.