A Quality perspective on the ethical questions behind AI in Pharma.

It all started from listening to a rather interesting artist who, as of today, also appear to be a successful entrepreneur, named Andrea Pezzi. A past as a TV presenter, a present as a businessman in AI.

It was his intervention at the Italian Senate held on 14 July 2026, his perspective regarding the entanglement between democracy, innovation and the impact of the digital revolution, that sparked much of my curiosity around this subject. Subject, no subjects. Singular, as all the meanings sit inside one big bubble.

Andrea Pezzi showed great concern about the relationship between technology, culture, knowledge and society. He approached AI not simply as a technological or economic issue, but as a question of power, responsibility, human judgement and the kind of society we want to build around it.

Are power, responsibility, judgment the same concern for us working in Quality in Pharma? And not just in relation to the introduction of AI.The speech made me realise that perhaps, when we talk about artificial intelligence, most of the discussion seems to concentrate on what we can see and measure:

• How we control it?
• What jobs may disappear?
• What AI will do to our businesses?
• Who owns the data?
• What legislation should be introduced?
• What our future will look like?

All of these are important questions. But I feel there are other angles that receives much less attention on what we are introducing:

• the ethical side (such as credibility, fairness and transparency) and
• the philosophical side (such as the concept of understanding versus the what is refer to be the Chinese Room Argument ).

Not only what can AI do? or how should we regulate it? but also:

• What should we allow it to do?
• What should remain a human responsibility?
• What happens to our judgement?
• Who will own the data?
• What will be able to do with that?
• Who will be remunerated for the “know-how” given to the servers?

Progress will not stop and I don’t think it should. At the same time, legislation will probably never move at the same speed as technological development. Regulation normally follows innovation rather than anticipating every possible consequence of it.

The pharmaceutical industry has another difficulty. AI is developing at a speed that does not sit particularly comfortably with some of the systems we have traditionally used to manage change.

I have already raised many questions that I do not intend to answer in this first post. In fact, that is probably the purpose of this blog. Not to provide final answers, but to raise questions that may be worth thinking about before the answers are decided for us. Some of them may prove uncomfortable. Some may turn out to be less important than they first appear. Others may become much more relevant as AI becomes more integrated into pharmaceutical development, manufacturing and Quality.

When, and I will, I revisit some of the issues raised, I am certain there will be more questions to answer.

 For now, I would leave the reader with one final question:

If AI becomes capable of making better and faster decisions than we do, how do we decide which decisions should still remain ours?

Cordially,

Daniele