Requirements and concepts for trustworthy procurement of LLM for digital forensics

An introduction to the Requirements and concepts for trustworthy procurement of LLM for digital forensics. Towards a reliable framework for determining Public procurement AI model risks.

While AI has revolutionised digital forensics by enhancing efficiency and accuracy in evidence analysis, there are concerns regarding the potential biases embedded in AI algorithms. These biases can impact decision-making processes and lead to unjust outcomes. Moreover, the reliance on AI may reduce human oversight and critical thinking in investigations, potentially overlooking nuanced details that human analysts could identify. Law enforcement agencies must continuously monitor and evaluate the performance of AI systems to ensure fairness and accuracy in forensic processes. 

By reducing bias from the outset of AI model procurement, this framework enhances investigation accuracy and efficiency, thereby strengthening the capabilities of forensic professionals within the judicial system.

This work looks at digital forensics model procurement and is working towards alignment against the AI ACT to help protect against AI's harmful impact on fundamental rights.

Presented at the 39th 2024 Bileta Conference at Dublin City University British and Irish Law Education and Technology Association conference.