Professor Musacchio, let's start with the definition of artificial intelligence?

Artificial intelligence is the ability of an automaton to manifest human characteristics such as reasoning, learning, planning, and even creativity. To date, we know three different types: limited artificial intelligence; General Artificial Intelligence and Super-Artificial Intelligence. The main risks associated with it concern the possible new forms of discrimination, the so-called "dehumanization" of the management of labor relations and the exercise of employers' powers, the risks of opacity of organizational and employer choices. Little or nothing is said about the risks associated with the use of artificial intelligence by criminals. I believe that this last aspect is a reality that we will soon have to deal with. We are, in my opinion, facing a transformation in the history of crime. In this sense, the crimes, the criminal reality, can no longer be the same. The specific question of new crimes and new methods of committing or omitting will therefore have to be read in this new and specific framework.

What new forms of crime could arise using artificial intelligence?

The use of artificial intelligence could, for example, act as a profit multiplier for money laundering activities, as it could achieve the commercialisation of products and the reuse of cleaned money on a large scale maximising reach and profits. On the net it is much easier for criminals to hide earnings. You can, for example, create digital wallets. Proxy mafias may be able to offer virtual banking services and quick access to financing and virtual currencies. They could offer criminal solutions to the problems that mafias have to solve on their own today. The modalities of illicit trafficking from drugs to trafficking in human beings and organs could also change. If for now we have only clues on these claims, soon I believe the evidence will also come.

In Italy we have recently seen many criticisms on ChatGpt, could thistool be able to create new forms of manifestation of crimes?

I believe that we will have to prepare ourselves for epochal transformations of organized crime, which can hardly be stopped if we do not equip ourselves in time. It is, however, a matter of being able to immediately understand the areas most at risk of these dangerous transformations in order to be able to adapt to similar developments and at least put a stop to them. For example, I believe that ChatGpt is also attractive to cyber criminals, who have already started using it to directly attack people and companies. There is, in fact, an objective risk on the cyber security front that could be exploited by organized crime to its advantage.

Can the new intelligent technology - which contains within itself the criminal evolution of manycenturies - then be reworked and adapted to our days?

If we look at what the inventors of artificial intelligence write, we soon realize that many more things can be achieved than we can imagine: from this point of view the criminal aspect is really "gruesome" and must worry a lot. "Crimes", "execution methods", "confidential information", can be reworked by a machine and can enter a criminogenic circuit of new evolution not easily predictable. Artificial intelligence, therefore, can act as a "database" to put them together in an "intelligent" way for criminal use by mafia organizations operating transnationally. Check Point Research researchers in Israel have identified at least three cases in which cybercriminals have used artificial intelligence for criminal purposes: 1) the creation of "stealer" software; 2) data encryption; 3) the creation of a dark web marketplace for exchanging data of stolen bank accounts and payment cards, malware tools, drugs, weapons and even human organs.

Do you not think, however, that the centrality of man is also essential in thissector?

Of course, the centrality of man as a criminal subject will remain. However, we must be aware of the criminal mechanisms that new technologies can create and consolidate over time. If artificial intelligence can be used for criminal purposes, however, it means that it can also be used to fight crime. It is on this last aspect that the States should focus.

You are a jurist and a criminologist, are you aware of court cases involvingmafias and the use of artificial intelligence?

At the moment I do not know of any judicial cases of organized crime using artificial intelligence in Europe, but in the field of money laundering something is moving in Latin America (Mexican and Colombian narcos are investing a lot of money on new technologies) and in Japan (the Yakuza has already experimented with a Machine Learning model), These are signs that we absolutely cannot underestimate. It is only a matter of time and we will discover that artificial intelligence is exploited by mafias 4.0 for multiple criminal purposes and so we will have the first judicial cases also in Europe and in Italy.

What can we do to prevent this from happening or at least to put at leastsome limits on it?

The first thing to be done as a matter of urgency is to equip the Member States of the European Union with the appropriate instruments to deal with the latest developments in organised crime in the field of technology. We will need to be able to effectively monitor the digital financial markets where major new crimes will be committed. The most modern mafias are already working to develop systems based on artificial intelligence for the manipulation of markets where the proceeds of crime can be laundered. Artificial intelligence can have "great advantages for mafiosi" precisely because they are "inundated" with money that they need to clean up in order to use it. The European Union is currently unable to counter defences to the extent of the misuse of artificial intelligence. The existing European bodies will have to be improved by filling the obvious gaps in the regulatory framework of the Member States that have so far allowed mafiosi to launder money using new technologies. We will categorically need a training course, law enforcement and the judiciary, which is capable of transmitting the new technological knowledge underlying the commission of new crimes resulting from artificial intelligence. We must immediately become aware of the sudden changes in the world of technology and of the ability of the new mafias to adapt quickly to these changes before it is too late.

Vincenzo Musacchio, forensic criminologist, jurist, associate at the Rutgers Institute on Anti-Corruption Studies (RIACS) inNewark (USA). He is an independent researcher and member of the High School of Strategic Studies on Organized Crime of theRoyal United Services Institute in London. In his career he was a pupil of Giuliano Vassalli, friend and collaborator of Antonino Caponnetto, an Italian magistrate known for having led the anti-mafia Pool with Falcone and Borsellino inthe second half of the eighties. He is one of the most accredited scholars of the new transnational mafias. Expert in strategies tocombat organized crime. Author of numerous essays and a monograph published in fifty-four States written withFranco Roberti entitled "The fight against the new mafias fought at transnational level". He is considered the greatest European experton the Albanian mafia and his in-depth work on the subject has also been used by legislative committeesat European level.