Antonio Teti: “The emergency is not artificial intelligence, but the defense of our cyberspace”

Professor Antonio Teti goes against the current regarding the risks of Artificial Intelligence, which seems to be the new global emergency after the alarms of Elon Musk, Geoffrey Hinton and the appeal of 350 “gurus” of the Center for Ai Safety. Teti is convinced that the problem is quite different, at least for Italy: lack of IT culture, skills drain, shortage of human and financial resources of the National Cybersecurity Agency. A counter-current voice but of specific weight: his penultimate book, Spycraft Revolution of 2021, bore the preface by the former head of the CIA in Italy, Robert Gorelick, and was praised by the general David Petraeus. Professor of Cyber ​​Security, IT Governance and Big Data at the G. D’Annunzio University of Chieti-Pescara, Teti does not hesitate to dampen the apocalyptic warning.

Why aren’t you afraid of Artificial Intelligence?

We are in a phase of machine learning, which allows Artificial Intelligence to learn from its mistakes but with a piloting that can always be traced back to the human being. It is false that Artificial Intelligence could become “sentient” in the coming years: with the development of the technologies in use, it is not possible to arrive at a platform that feels emotions. Predicting what will happen in fifty years is a monstrous gamble. Science fiction. One can philosophize about apocalyptic scenarios, but there are no scientific elements that prove the fears.

Wouldn’t regulation be appropriate?

It is bluster to think of regulating the development of technologies. There is no world government that can bind all the countries of the planet. Even if an agreement were to materialize between some, there would be others who would evolve their own research. The same goes for disinformation, which can have shocking effects. Anyone who thinks they are regulating cyberspace is a madman or a utopian: a rule that is valid in Italy or in the European Union cannot be imposed on the world. And then on social media managed by private companies, the States can do little. What was the ChatGPT block in Italy for, if a vpn was enough to get around it? Every national standard is technically useless.

An example?

China: since 2017, a law has legitimized spying abroad for any Chinese citizen: if he works in a foreign company and obtains information useful to his government, that activity is a crime for us but he is rewarded with family benefits.

What has changed with the conflict in Ukraine?

It has made the risks more evident: on Twitter there are profiles that have gone from a few hundred followers to 800,000 in a few weeks and they do not belong to geopolitical experts but to improvised ones, assuming they are real people and not fake accounts for disinformation purposes. On platforms, users are also suppliers of information: the attractiveness of a profile counts. If it has a polarizing language, and produces captivating audiovisual content, it increases its credibility exponentially.

How do we defend ourselves?

With training on the correct use of technologies. The hacking of the L’Aquila Local Health Authority demonstrated that adequate safety tools and methodologies had not been adopted. We are talking about sensitive data, which are far more important than personal data: how much is it worth for a pharmaceutical company to get their hands on it?

Do you need institutional training?

We need more IT culture in general. We need to start from elementary school: kids tinker with devices without even considering their effects. The pandemic has produced an acceleration in the use of digital tools, but at the same time negative consequences with a lowering of the level of students, because distance learning can only be used on an ancillary basis: learning in the classroom without interference is something else.. And in the world of work, smart working, even if it is convenient for some companies because it reduces costs, is producing more worrying behavioral changes than Artificial Intelligence. Staying at home is comfortable, but with deadly risks.

How is Italian cyberspace defended?

The Pnrr allocates around 600 million to cybersecurity: they seem like a lot but they are nothing. The Cybersecurity Agency was born in 2021 with a twelve-year delay on similar French and German institutions, which enjoy much larger resources and personnel that reach 2,500 operators against our current 170. With these numbers, we cannot defend the Village. We need more funds and we need to speed up recruitment with adequate salaries: computer engineering graduates are going abroad. Public and private companies must understand that it is not enough to buy firewalls and antiviruses if they don’t know how to use them. It’s like having the best rifles without training soldiers.

Who threatens us in cyberspace?

Only in a few cases can the precise paternity be traced back. Cyber ​​gangs act as mercenary enforcers for different purposes ranging from blackmail to espionage. Identifying the director is difficult. And then it’s not always a question of “enemies”, the attack can come from an allied country that intends to push for certain decisions. To understand this, it is also necessary to have an Open Source information collection in cyberspace that only Artificial Intelligence allows, with a scraping activity to be made available to the analyst. How do politicians to identify the moods of the population.