The rights of individuals in the criminal justice system could be infringed by the use of new technologies in policing, the Faculty of Advocates has warned.
Faculty was responding to the call for evidence from the independent advisory group on emerging technologies in policing.
Its response notes that concerns raised around the use of technology are not new, but the situation is being exacerbated as technologies in use became more sophisticated and more widely deployed, such as AI (artificial intelligence) systems.
"The process of detecting crime and collecting evidence may often involve the use of technological means. These means are becoming more sophisticated, more efficient and more intrusive than in the past. This improving of technology and more extensive deployment thereof (for example, the extensive use of CCTV cameras and number plate recognition systems) potentially presents challenges to fundamental rights, consequently bringing a need to ensure that such deployment is proportionate and otherwise complies with the European Convention on Human Rights", it states.
Examples already exist of where AI can be deemed to have infringed on the rights of individuals in the justice system, the response adds. "One such example was the use of a proprietary artificial intelligence system which purported to be able to predict the likelihood of a particular offender committing further offences if released on bail in the state of Wisconsin. The decision of the Wisconsin Supreme Court, which approved the use of that system, has since come in for heavy and sustained criticism."
Faculty further believes that the increasing use of biometric technologies, such as facial recognition systems, also raises concerns, particularly where these are combined with AI systems that analyse biometric data. "The issues this raises become even more acute due to the real risks that the datasets used to educate the AI system will contain inherent biases."
The use of biometrics for "affect recognition" is of particular concern. A police force in China is reported to be aiming to use photographs of drivers taken from roadside cameras to determine from their facial appearance whether to stop them for drunk driving, and there are systems that claim to be able to predict a person's criminality. While some of these claims are bogus, given that they are presented under the guise of AI there is a danger that they could be viewed as being fact-based.
"Therefore, in analysing the adequacy of present legal controls and protections, there is a need to be ever mindful of the difference between conventional developing technologies and those which involve the use of AI systems,” Faculty continues.
"It will be appreciated", it concludes, "that proper oversight and scrutiny is essential. The Faculty has no particular view on the mechanism which should be used to guarantee that oversight and scrutiny, but does consider it critical that legal safeguards are in place to ensure that any use of new technologies which may infringe an individual’s fundamental rights remains within the law."