Artificial intelligence to anticipate cyberattacks – launch of the Cybi startup

Information systems, connected objects, industrial systems… these environments face increasingly frequent and sophisticated computer attacks. Companies are prime targets for hackers and must protect themselves from this growing threat. The new Cybi startup is the fruit of the cybersecurity expertise of the RESIST team at the Loria laboratory (CNRS, Inria, Université de Lorraine). It provides offers solutions for analyzing computer attack paths and intelligently automating cybersecurity operations. Cybi has just been launched in Nancy by the Inria and the Université de Lorraine with support from the Lorrain Incubator and the Grand Est region.

State-of-the-art technology

The majority of cybersecurity solutions analyze individual vulnerabilities in computer systems. However, hackers are increasingly operating in attack paths using critical vulnerabilities and also other flaws as entry points to penetrate the computer network and infect the entire system. A connected object (connected socket, internet box, industrial automaton, etc.) with a good individual security score can thus become a crossing point in an attack path to reach a critical target when it is networked with other objects.

This breakthrough technology developed by Cybi’s team of researchers and engineers was patented in 2020. It relies on artificial intelligence models and algorithms to analyse all the attack paths present on a network and prioritize the vulnerabilities that need to be corrected, thus reducing risks and incident response times.

Detecting and prioritising vulnerabilities thanks to artificial intelligence

Scuba relies on corpora of vulnerabilities. When a flaw is discovered, during pentesting (methods for evaluating the security of a system) or bug bounty (a reward programme for bug hunting) operations, manufacturers must publish an open-access patch. This information is notably available in CVE (common vulnerabilities exposure) databases.

“Our tool is able to easily read and understand these large amounts of text thanks to automatic language processing and artificial intelligence methods. It can thus automatically analyze vulnerability reports and detect potential attack chains“, explain the researchers. “We are also working on a test platform dedicated to studying the vulnerabilities of connected objects so we can assess product security levels“.

The team has also recently tested its technology on well-known viruses such as Pegasus and Puzzle-maker and was able to identify their attack paths.

Cybi, an essential ally for business security

The start-up is now ready to work with cybersecurity experts, large companies and more particularly with Security Operation Centres (SOCs) which guarantee information security. Scuba appears to be an essential link in the system security chain.

“Thanks to our vulnerability and vulnerability chain analysis solutions, Cybi offers a real decision support tool, an analysis of real risks and personalized recommendations to implement the best techniques to correct flaws”, specifies Régis Lhoste, president of the startup.

“The team’s expertise and maturity in cybersecurity position Cybi as an innovative and visionary startup. Our Scuba solution arrives at just the right time on the cybersecurity market. Scuba offers advanced features that are in high demand and have been not been used enough until now to respond to the threats of cyberattacks that are increasing daily. Our various exchanges with the main actors in French cybersecurity are very positive. Scuba is THE solution all the structures involved have been waiting for, whatever their profile or size“, points out Fabian Osmond, general manager of the startup.