'Artificial intelligence alone is not enough: skills, governance and culture are needed'
The analysis by Chiara Sgarbossa, head of the Politecnico di Milano's Digital Health Observatory, between opportunities and grey areas
2' min read
2' min read
Artificial intelligence has entered the drug life cycle and the doctor-patient relationship. But the real transformation requires more than technology. Chiara Sgarbossa, director of the Milan Polytechnic's Life Science Innovation Observatory, guides us through the numbers, trends and criticalities of the sector.
Let's start with generative Ai. What solutions do Italian pharmaceutical companies use? .
Most use generalist generative Ai (60%), while 36% have already developed ad hoc solutions, using their own data. This is a good sign, but there is a need to increase the use of corporate tools in order to have the most reliable answers possible, protect corporate know-how and create competitive advantage.
Are the current standards sufficient?
Today, the Ai Act is the main reference regulation. The Ema has published a reflection paper, but these are guidelines, not obligations. There is a need to raise awareness within companies, especially about sensitive data and the responsibilities of using AI in work activities.


