Police Scotland must adhere to 200-year-old Peelian Principles when considering AI surveillance tech

CCTV cameras can be used by police to search for people on a watchlist using AI-powered live facial recognition technology (Picture: Leon Neal) | AFP via Getty Images

As Police Scotland considers using live facial recognition technology, problems experienced in other parts of the UK where it has already been introduced and the effect on the idea of ‘policing by consent’ must be carefully considered

The concept of “policing by consent” is a historic one. The so-called ‘Peelian Principles’ – named after Prime Minister Robert Peel and first adopted in 1829 – included a recognition that the “power of the police to fulfil their functions and duties is dependent on public approval of their existence, actions and behaviour and on their ability to secure and maintain public respect”.

Given the approach of its 200th anniversary, this idea has more than stood the test of time, with policing by consent now regarded as “a fundamental principle that underpins modern democratic societies”, to quote the UK Justice Ministry.

However, this does not mean that police forces should remain wedded to the past. Indeed, they are necessarily required to continually modernise their methods if they are to stay one step ahead of the criminals.

Police Scotland’s decision to think about introducing live facial recognition technology – in which artificial intelligence is used to scan live camera feeds for people on a ‘watchlist’ – is, therefore, something they would be remiss not to consider.

Police Scotland urged to abandon facial recognition tech plans amid warnings it is unethical

However, with policing by consent in mind, they need to look at problems that have emerged in other areas where it is already being used. For example, live facial recognition technology has helped South Wales Police make 72 arrests, but the system produced 2,833 false alerts. The courts also found the way it was used initially breached privacy rights and equalities law.

In 2019, a report found the Metropolitan Police’s AI facial-recognition bot was correct on just 19 per cent of occasions. There appears to be a particular problem with recognising the faces of people from ethnic minorities. For example, Shaun Thompson, a black anti-knife crime activist, was threatened with arrest because of one of the system’s many mistakes.

There is almost certainly a place for AI and facial recognition technology in policing. However, there are clearly also human rights and operational questions that need to be answered satisfactorily.

A failure to do so would risk a breach of a tradition that has served this country well for two centuries.