- Status: Project phase
Background
IT forensics describes the securing and analysis of criminally relevant data from IT systems using scientific methods. Strictly methodical data analysis is used to prove facts in court. Digital forensics also plays a decisive role in the application context of artificial intelligence, both for AI systems in private use and for frequently networked AI systems in companies. The understanding of when and why evolving and self-overwriting AI-based techniques fail must therefore be expanded and improved by analyzing them with forensic methodologies and tools.
Aim
The research program “Forensics of Intelligent Systems” addresses the expansion or research of existing or fundamentally new forensic procedures and tools, taking into account legal standards. The aim is to provide proactive “forensic readiness” and “forensic-by-design” methods for the legal detection of tampering in evolving, self-overriding AI systems. These novel and method-oriented research approaches for the legally compliant proof of misconduct must be transparent, fair and explainable. The methods are to be evaluated and improved using inversion and evasion attacks as well as data poisoning attacks on an AI system. The research areas addressed within the program include ‘Intelligent Text Processing’, ‘Intelligent Image and Video Processing’ and ‘Intelligent Autonomous Systems’.
Disruptive Risk Research
Within digital forensics, a lot of research is currently being conducted into how AI systems can support internal and external criminal prosecution, e.g. when writing forensic reports. There are hardly any concrete research projects that make the (mis)behavior of artificial intelligence forensically traceable in accordance with legal regulations. The program therefore has an innovative and disruptive character. For this reason, one contractor was selected for each of the three research areas following an extensive review and intensive rounds of negotiations.
- Project ‘Intelligent Security Inspection of Data using forensic Extraction and Analysis(INSIDE)’ at Mannheim University of Applied Sciences
- Project ‘Development of forensic methods for legally compliant auditing of systems with artificial intelligence(EMAS-KI)’ of Atos Information Technology GmbH
- Project ‘Forensic methods of AI systems in the mobility industry (FORAImobility)’ of ZF Friedrichshafen AG