Rhys Hibbert couldn’t walk across a room without glasses and a stick. A tumor on his pituitary gland was pressing on the nerves that control his sight, and it was slowly stealing his vision. Then a team of London neurosurgeons went in to remove it, and they brought an AI tool along for the operation. According to Futurism AI, that surgery saved his eyesight, and when Hibbert woke up he could see everything in the room clearly.
Here’s what actually happened.
What went down
The operation took place at the National Hospital for Neurology and Neurosurgery in May, though details were kept quiet until now, Futurism AI reports. Surgeons removed a large benign tumor from Hibbert’s pituitary gland, a spot that sits at the base of the brain, hidden from view and packed in next to the arteries feeding the brain and the optic nerves running to the eyes.
The AI didn’t hold a scalpel. It watched. A camera fed through the patient’s nose to the base of the skull streamed live footage, and the tool analyzed it in real time. It tracked the surgeons’ instruments, flagged where hidden blood vessels and nerves were likely to be, and highlighted the safe zones to cut. The BBC compared it to facial recognition, except for anatomy you can’t see.
Why a millimeter matters here
The pituitary gland is one of the least forgiving places in the body to operate. Health officials told The Guardian that going a millimeter wrong can be the difference between success and blindness, stroke, or death. As Hibbert put it to the BBC: “There are no road signs inside our head.”
That’s the real value of the tool. It was trained on hundreds of surgical videos of pituitary tumor removals, with researchers manually drawing around every vessel and nerve so the system learned exactly where they hide. Sophia Bano, an associate professor in robotics and AI at University College London and the technical lead on the project, told The Guardian the AI “has been exposed to a breadth of surgical examples that would take a surgeon many years to encounter.”
That’s the part that stands out to me. A surgeon builds pattern recognition over a career, one case at a time. This system ate hundreds of cases before it ever entered an operating room.
Why this matters for the industry
Most medical AI so far has lived at the edges of care. It writes clinical notes, helps look up symptoms, reads scans before a doctor confirms. This is different. This is AI inside a live, high-stakes operation, guiding hands in real time on one of the most delicate procedures there is.
The humans stayed in control the whole way through. That’s the key line, and it’s the model everyone’s watching: AI as a co-pilot, not the pilot. When the margin for error is a single millimeter, a second set of eyes that never blinks is a genuine safety upgrade.
The uncomfortable part
An incredible result still leaves open questions. What happens when the tech gets it wrong? Who’s accountable when a machine flags the wrong safe zone?
There’s a quieter worry too. Experts told the BBC that surgeons might lean on these tools too hard, especially a new generation of med students trained on AI from day one. Normally a surgeon studies a patient’s scans until they know that specific anatomy cold before cutting. If the AI becomes a crutch, that hard-won intuition could fade.
What to watch next
Expect more of this, fast. Once a tool proves it can save a patient’s sight in a procedure this unforgiving, hospitals and device makers will push to scale it to other operations. The debate ahead won’t be whether AI belongs in the operating room. It’ll be about where the line sits between assisting a surgeon and replacing their judgment.
For Hibbert, the verdict is already in. “It’s given me my life back,” he said.
Full details are available at the original Futurism AI report.