Using artificial intelligence to reduce the effect of human biases in forensic science [Video]

Any scientific process used as part of a criminal investigation is considered forensic science, and many of the methods involved are visual, meaning they rely on the interpretation and expertise of the observer. This introduces cognitive biases into the investigative process, potentially affecting the overall decisions made. Sherry Nakhaeizadeh and her colleagues at the UCL Centre for the Forensic Sciences, UK, are currently working on the development and application of new artificial intelligence technologies and approaches for improving forensic anthropology methods.

These technologies are being used in complex visual tasks, such as employing eye-trackers to study gaze pattern strategies and decision-making processes involved in the assessment of skeletal remains and crime scene examinations. “If we get the human and machine to work together, we can address some of the challenges that forensic science is currently facing,” says Nakhaeizadeh.