We are delighted to announce that Martin Isztin, under the co-supervision of Gábor Antal PhD and Péter Hegedűs PhD from FrontEndART Ltd., has earned significant recognition at the 2024 Spring Scientific Student Conference IT Section, held at the University of Szeged, Hungary.

Martin secured 1st place with his outstanding research titled “Investigation of Automatic Generation of Vulnerability-Witnessing Unit Tests,” conducted as part of the Sec4AI4Sec project. His research focused on employing GPT, one of the most prominent large language models to automatically generate unit tests that detect software vulnerabilities. The results demonstrated that GPT generated syntactically correct test cases 50% of the time. Although only 15% of these cases were semantically correct, the tests required minimal adjustments to become fully functional.

This research was carried out under the supervision of FEA colleagues and addresses a research problem within the framework of the Sec4AI4Sec project.