Why this webinar is a must-watch for forensic and expert witness professionals.
The Forensic & Expert Witness Community recently hosted a thought-provoking webinar exploring the evolving role of artificial intelligence in forensic investigations and expert witness work. Chaired by Phil Southall, Chair of the Forensic & Expert Witness Advisory Group, the session featured expert insights from Steve Molloy, Principal at Accuracy, and Calum Mackenzie, Senior Optimisation Specialist at Accuracy.
This webinar offered a balanced, critical look at how AI is reshaping the forensic landscape—what’s working, what’s worrying, and what professionals need to prepare for.
From Hype to Integration: Where AI stands today
The discussion began with a recap of predictions made at the 2024 Annual Conference. Steve and Calum revisited their forecast that AI would move from hype to practical integration and they were largely right. AI tools, particularly large language models (LLMs), are now embedded in workflows across forensic investigations, from document review to data analysis.
But the speakers were quick to point out that this isn’t a story of seamless transformation. While AI has made certain tasks more efficient like first-pass document reviews it hasn’t replaced the need for human expertise. Instead, it’s shifting the focus toward second-level reviews, where subject matter experts play a more prominent role.
The discussion also highlighted the importance of building robust workflows around AI tools. LLMs are powerful, but they require precise prompts and structured processes to be effective. Broad or vague questions won’t yield reliable results, and the technology is far from being able to autonomously detect fraud or anomalies without human guidance.
Risks, Realities, and the Road Ahead
One of the most compelling aspects of the webinar was its candid exploration of AI’s limitations and risks. The speakers tackled issues such as:
- Verification challenges: AI can generate plausible but incorrect outputs, making it risky to rely on without thorough human oversight.
- Cognitive impact: A recent study suggests that heavy reliance on AI may reduce critical thinking and knowledge retention.
- Manipulation risks: AI systems can be gamed by bad actors, raising concerns about the integrity of forensic evidence.
- Bias and truthfulness: LLMs don’t understand truth—they generate responses based on patterns, not facts, which is problematic in high-stakes environments.
These concerns aren’t just theoretical. They have real implications for how forensic professionals work, how courts view AI-generated evidence, and how junior staff are trained in an AI-assisted world.
Steve and Calum also touched on broader industry trends, including the tension between open and closed AI models, data sovereignty, and the legal and ethical questions surrounding AI training data. As Mackenzie noted, we’re entering a phase where our relationship with technology itself is being redefined.
Watch On-Demand
This webinar is essential viewing for anyone in the forensic and expert witness space. It doesn’t just inform it challenges assumptions, sparks debate, and encourages professionals to think critically about how they use AI in their work.
Whether you're optimistic, indifferent, or sceptical about AI’s impact, this session offers valuable perspectives that will help you navigate the evolving landscape with confidence and caution.
A huge thank you to Steve and Calum for their insightful and honest contributions, and to Phil Southall for expertly guiding the discussion.
If you missed the live session, we strongly encourage you to watch it on demand. The conversation is timely, relevant, and packed with practical takeaways for forensic practitioners at all levels.
