Informace o projektu

Anotace je dostupná pouze v anglickém jazyce.

With the rising threat of crime in cyberspace, a growing need emerges for forensic investigation of these cybercrimes. Such investigations are highly time-consuming with delicate supportive processes. The problem of systematic preparation for the potential forensic investigation during the software development process, called forensic readiness, has only been explored recently. Therefore there are still many open issues and challenges, with missing methods and guidelines that would support software engineers in building forensic-ready software systems.

One of the essential open challenges in the cybersecurity and forensic-readiness space is the risk of malicious intentions of smart software agents joining a software ecosystem (e.g. AI-enhanced software update downloaded to an autonomous driving car). To understand the intentions of such a smart software agent and decide whether it can be trusted and given access to critical system functions, forensic-driven evidence collection can be of enormous help to detect malicious intentions or trust attacks that the smart software agent might involve in to improve its trustability or whitewash its reputation.

The project aims to create methodological support for the development of AI-enhanced software ecosystems prepared for real-time and post-incident forensic analysis, which in the event of a cyber-security incident would be able to reveal the circumstances of the incident, and point to the accountability of the individual software agents. The method will be centered around the concept of trust management and forensic-driven evidence collection. The forensic-analysis support will be designed to produce complete, undeniable and authentic information forming digital evidence with potentially high evidentiary value, to feed back to the trust management process.