His research shows that many serious digital incidents arise because sensitive or malicious data moves unnoticed through different parts of software and only causes harm after multiple steps.
Iather than viewing a cyberattack as one large incident, Johannesmeyer demonstrates that attacks are better understood as a chain reaction of small, seemingly harmless steps. By tracking and analyzing these steps individually, vulnerabilities can be discovered that traditional security methods often miss. Using this approach, thousands of previously unknown security flaws were identified in widely used software, including operating systems that form the backbone of servers and smartphones worldwide.
Fewer false alarms with this method
One key result is that this method generates far fewer false alarms than existing security systems. This makes the technique not only more accurate, but also more practical for developers and security teams. They can focus more effectively on real risks without being overwhelmed by alerts that ultimately turn out to be harmless.
Critical digital systems -such as those used in hospitals, energy infrastructure, cloud services, and personal devices- are becoming increasingly complex and interconnected. In such systems, small mistakes in handling data can have major consequences, ranging from data breaches to disruptions of essential infrastructure. By gaining better insight into how data flows through software, these risks can be identified and prevented at an earlier stage.
In practice, this means that security problems do not only come to light after an incident, but can already be detected during the development process. Developers can automatically test whether unauthorized or untrusted data could influence critical decisions within their software. In this way, the research contributes to more secure digital systems and reduces the risk of societal disruption caused by cyberattacks.
Johannesmeyer defends his PhD Januari 12 at Vrije Universiteit Amsterdam.