Experts Say AI Can Boost Forensic Probes With Strict Checks
Lucknow, October 23: Experts have said that artificial intelligence (AI) has the potential to make forensic investigations quicker, more efficient and more precise. The technology can assist investigators in areas ranging from fingerprint and DNA analysis to CCTV footage review, digital evidence examination and mapping relationships between suspects and incidents.
But the experts cautioned that accepting AI outputs without verification could be risky for both investigations and court proceedings.
The remarks came during a panel discussion titled 'Artificial Intelligence in Forensics and Investigation' at the Uttar Pradesh State Institute of Forensic Science in Lucknow. The event was held as part of the 'TechSangam' segment of Samvad-26, a three-day international dialogue. Speakers from within the country and abroad took part in the session on Friday.
The panellists agreed that AI is a helpful auxiliary tool in criminal investigations. However, they stressed that it cannot substitute the expertise of professionals, human discretion or the independent corroboration of evidence.
Moderating the session, DIG (Technical) Ashish Tiwari welcomed the speakers and delegates. He underlined the importance of forensic science in preventing, identifying, investigating and prosecuting crime. He also referred to the Bharatiya Nyaya Sanhita (BNS) and the broader role of forensic evidence in the criminal justice system.
The central conclusion of the panel was that AI can serve as an effective technical resource for investigating agencies. Its value, however, will rest on dependable data, transparent procedures and consistent human supervision. Experts noted that as technology accelerates, the authenticity of evidence and the credibility of judicial processes must be preserved.
Sandeep Joshi explained that AI can process large volumes of data in a short span, speeding up investigative work. Information that usually takes long to review manually can be scanned for patterns and probable links, giving investigators timely leads and helping them make informed decisions.
He pointed out several applications, including matching fingerprint patterns, analysing complex DNA datasets, spotting suspicious activity in CCTV recordings and tracking the movement of people across multiple locations. AI can also sift through extensive forensic records to identify potential connections between individuals, events, places and items of evidence.
According to Joshi, the performance of AI depends on the quality of data, the appropriateness of software and hardware, and the method of analysis. He said reliable results require attention not only to the technology itself but also to the processes governing its use.
The panel also debated questions about AI reliability. Experts observed that while AI can strengthen investigations, flawed or prejudiced data, misleading outputs and manipulation of digital evidence can create major difficulties. AI may at times generate conclusions that lack adequate support from the available material. Systems trained on skewed data can distort results, and deepfakes and other synthetic content raise the danger of fabricated evidence.
To counter these risks, experts recommended independent confirmation of AI findings, verification of the authenticity of original evidence and thorough documentation of the entire analytical process. Investigators were advised to cross-check AI outputs against other trustworthy sources and conventional forensic techniques. They also stressed human supervision, accountability and well-defined legal and ethical norms.
A recurring theme was the idea of 'AI for security and security of AI'. Experts said the benefits of AI in investigations will materialise only when systems are shielded from errors, interference, misuse and unauthorised access.