AI instead of a doctor prescribes drugs and makes consultations: from Utah to El Salvador

Artificial intelligence is already a reality in everyday clinical activity, for example, it is widely used in diagnostic equipment such as MRIs and CT scans of the latest generation, flanking and helping white coats to better read the results of examinations. But so far, it has not yet taken the place of doctors, as it is beginning to do in Utah in the United States or in El Salvador, where artificial intelligence can not only prescribe medicines to patients but also provide initial medical advice. Mind you, in these two maxi-trials, Ai has not abolished the figure of doctors, but they are at the same time two experiments that open the door to an as yet unexplored world that can on the one hand speed up and simplify life for patients, but on the other hand loosen the relationship between doctor and patient that is already very frayed today.

The trial in the state of Utah started last January, the first of its kind, and will last for a year. It is the result of a collaboration with a company called Doctronic, which has released a tool based precisely on artificial intelligence that makes it possible to make pharmaceutical prescriptions without the involvement of doctors. The software has so far only been programmed for renewing prescriptions for medicines that patients periodically take for already diagnosed conditions, and specifically covers less than 200 drugs including corticosteroids, statins, antidepressants, hormones and anticoagulants. For the time being, prescriptions for opioids, injectable drugs and therapies for Adhd are excluded. The trial has also been at the centre of an article in Jama in recent days that tried to line up the benefits such as the reduction in errors - the tool has shown an accuracy of over 99% - and greater access to treatment, but also risks such as algorithmic opacity and above all weakening of the doctor-patient relationship.

In recent days, the president of El Salvador,Nayib Bukele announced that he will entrust much of the management of the public health system to Google's artificial intelligence. In a national radio and television address, the conservative leader explained that his countrymen will have access to a mobile application called Doctorsvapp, where it will be possible, among other novelties, to get an initial consultation and order laboratory tests, but also to monitor the treatment of patients suffering from chronic diseases such as diabetes, hypertension and high cholesterol. This is the second phase of the smartphone programme launched in November 2025 and supported by Google's Gemini AI. "I am very excited, because we are creating the best health system in the world," Bukele said during the event that was also broadcast on social media. The announcement was made together with Guy Nae, Google Cloud director for the Public Sector in Latin America, and specialists Edgardo Von Euw and Stella Aslibekyan. "The spirit of this programme is that no Salvadoran suffering from a chronic illness should have to give up treatment," said Euw.

And in Italia? Pnrr funds have been used to finance a project that has created a platform for family doctors that can be queried from a PC - also in the form of an app - by means ofa user-friendly conversational interface that includes a technological architecture based on artificial intelligence. The doctor will be able to ask questions to 'Mia', which in turn will draw on a 'validated and certified' knowledge base: these are information sources and official documents such as guidelines, protocols, regulations, scientific literature, diagnostic and therapeutic care paths, nomenclators, etc. The project, which started recently, envisages a progressive implementation by the first 1500 family doctors experimenting with the three cases. During this year, a training plan will also be implemented to enable medics to use the tool and increase their knowledge of artificial intelligence. Finally, the last phase envisaged is that of the management and maintenance of the platform, starting in January 2027, which envisages a gradual extension to the other family doctors working for the SSN: after the first 1,500, it will be extended to another 3,000 white coats in 2027, then another 7,500 and finally another 15,000.