Cybersecurity is in transition as new technologies emerge

Workshop brought together Brazilian and American experts to discuss new opportunities for collaborative research in the face of increasing volumes of information, the rise of artificial intelligence, the imminent arrival of quantum computing, and the mass production of fake news.

By José Tadeu Arantes  |  Agência FAPESP – With the expansion of cyber threats, the protection of digital infrastructures has become a major global challenge. In this context, Brazilian and American researchers met in an immersive workshop to discuss advances in cybersecurity and seek scientific collaborations. The event, “NSF-FAPESP Workshop on Cybersecurity and Privacy,” promoted by the U.S. National Science Foundation (NSF) and FAPESP, focused on three main areas: network security; artificial intelligence and security; and information ecosystem, security of veracity and origin.

“Network security is a consolidated field, but it faces growing challenges with the increase in the volume of information and the emergence of new technologies, such as 5G networks, quantum communication and the rise of artificial intelligence,” says Marcos Simplício, professor at the Engineering School of the University of São Paulo (POLI-USP) and coordinator of the event.

Simplício points out that encryption protocols, advanced firewalls, and intrusion detection and prevention systems are fundamental to ensuring the reliability of digital infrastructures. It is worth remembering that firewalls are security devices that control traffic in and out of networks or computers. “The zero trust approach has been gaining ground because it’s based on the principle that no entity should be trusted as a matter of principle, regardless of its location in the network,” the researcher comments. 

The goal of the workshop was to foster connections between Brazilian and American researchers, to identify common interests, and to create opportunities for collaborative research capable of producing innovations and making a difference in their respective fields.

“Often, what happens is that the research institutions propose the problems and the researchers chase the solutions. In this case, we decided to turn the process on its head, so that the researchers themselves could point out to the institutions what the fundamentally important problems are in the areas in which they work. The procedure we followed – which I’d already adopted at other events and which we decided to adopt at this one – was what we call the Blue Sky Approach. It basically consists of encouraging participants to enter into an open discussion and raise a large number of questions. From this initial round, we make a selection and identify what everyone really agrees on,” said Marco Carvalho, executive director of the Harris Institute for Assured Information (HIAI) at Florida Tech (Florida Institute of Technology) and co-organizer of the workshop.

The event brought together 28 Brazilian researchers and 26 U.S. researchers. After four days of discussions, they selected eight or nine collaborative research proposals with their respective teams to submit to the NSF and FAPESP.

Artificial intelligence was, of course, a hot topic. As the technology advances and becomes more widely used, it has become both a resource for digital security and a target for threats. Generative AI presents risks such as the manipulation of responses through adversarial attacks, hallucinations that generate false information, and contaminated training that can introduce biases into models. On the other hand, “AI can be used to improve cybersecurity by enabling the analysis of suspicious behavior and the automation of incident response,” says Ricardo Dahab, full professor at the Institute of Computing of the State University of Campinas (UNICAMP) and a specialist in cryptography. 
 
Dahab points out that one of the most important recent changes is the use of AI to detect attack patterns and protect critical infrastructure, such as communications networks, financial systems, and large public or private institutions. However, he warns that the same technology can also be used to facilitate attacks, increasing the sophistication of the threats. “We’re moving away from just looking at systems and their vulnerabilities to observing behaviors and patterns in order to anticipate cyber threats,” he adds. 

Still on the subject of artificial intelligence, Carvalho adds that the principles being used are not new. They’ve been around for a long time, since the 1950s or so. What has exploded is the processing power to do the matrix calculations required for this type of application.

“AI uses huge vector spaces. And what made this possible was the gaming industry, which created processors that could do the tasks required. So models that would’ve been limited to very specific applications are being popularized for anyone. This is both good and bad. It’s very good because the more people use it, the more artificial intelligence improves and the more it helps people solve problems. Much of what we do today is already being mediated to some extent by AI. On the other hand, it’s not a finished technology. It still has major limitations. We need to be very aware of that. What researchers are doing is basically monitoring this process and adjusting the capabilities of artificial intelligence to get the best out of its use,” Carvalho explains.

Another major cybersecurity challenge is the emergence of quantum computers. As has been widely reported and discussed, quantum computers can break most current encryption systems due to their ability to perform calculations exponentially faster than classical computers. “The field of cryptography is going through a phase of transition to new methods, new algorithms, because of the advent of quantum computers. So the community is already working hard on a new generation of algorithms, which should make up the so-called post-quantum cryptography,” Dahab points out. 

Brazil has the intellectual capacity to take off in cybersecurity. According to Simplício, the main cybersecurity research centers in Brazil today are USP and UNICAMP, with relevant work in network security and cryptography, as well as the Federal University of Pernambuco (UFPE) and the Center for Advanced Studies and Systems of Recife (CESAR), which focus on malware (malicious software created to infect, damage or gain unauthorized access to devices and networks) and cyber-attacks. “Initiatives such as the National Data Protection Authority [ANPD], the National Cyber Defense Committee [CNDCyber] and the Hackers do Bem program are examples of positive cybersecurity actions taking place in Brazil. One of the major challenges is to guarantee the privacy of users while at the same time enabling the investigation and prevention of cybercrime,” says Simplício.