Public attitudes towards the use of automatic facial recognition technology in criminal justice systems around the world

Abstract

Automatic facial recognition technology (AFR) is increasingly used in criminal justice systems around the world, yet to date there has not been an international survey of public attitudes toward its use. In Study 1, we ran focus groups in the UK, Australia and China (countries at different stages of adopting AFR) and in Study 2 we collected data from over 3,000 participants in the UK, Australia and the USA using a questionnaire investigating attitudes towards AFR use in criminal justice systems. Our results showed that although overall participants were aligned in their attitudes and reasoning behind them, there were some key differences across countries. People in the USA were more accepting of tracking citizens, more accepting of private companies' use of AFR, and less trusting of the police using AFR than people in the UK and Australia. Our results showed that support for the use of AFR depends greatly on what the technology is used for and who it is used by. We recommend vendors and users do more to explain AFR use, including details around accuracy and data protection. We also recommend that governments should set legal boundaries around the use of AFR in investigative and criminal justice settings.

Fig 1. Focus group overarching themes.

Graphical representation of overarching themes and themes identified from focus groups conducted in the UK, Australia and China.

Fig 2. Society overarching theme.

Graphical representation of Society overarching theme and its component themes and subthemes identified from focus groups conducted in the UK, Australia and China.

Fig 3. Technology overarching theme.

Graphical representation of Technology overarching theme and its component themes and subthemes identified from focus groups conducted in the UK, Australia and China.

Fig 4. Purpose overarching theme.

Graphical representation of Purpose overarching theme and its component themes and subthemes identified from focus groups conducted in the UK, Australia and China.

Fig 5. Question 4 data.

Responses to question 4 “How aware are you of the use and adoption of facial recognition systems in your country?”.

Fig 6. Questions 5–7 data.

Percent of participants who agreed (responding 4, 5 or 6 on the scale of agreement) to questions 5–7 “To what extent do you agree with facial recognition technology being used by the police/the government/private companies in your country?” for each use case. A, police. B, government. C, private companies.

Fig 7. Questions 8–9 data.

Percent of participants who were comfortable (responding 4, 5 or 6 on the scale of comfort) to questions 8 and 9 “How comfortable do you feel with police in your country using facial recognition technology to search for individuals who are/are not on a watchlist”.

Fig 8. Questions 10–12 data.

Percent of participants who trust (responding 4, 5 or 6 on the scale of trust) these users to use AFR responsibly, responding to the questions 10–12 “To what extent do you trust the police/the government/private companies in your country to use facial recognition technology responsibly”.

Fig 9. Questions 13–15 data.

Percent of participants who agree (responding 4, 5 or 6 on the scale of agreement) with each use case in court (questions 13–15).

Fig 10. Question 16–17 data.

Percent respondents responding accurate (4, 5 or 6 on the scale of accuracy) to question 16 “How accurate do you think this technology is at identifying the correct person from a database?” and 17 “How accurate do you think this technology is at recognising the same person across changes in their appearance?”.

Fig 11. Question 18 data.

Responses to question 18 “How accurate do you think this type of facial recognition technology is compared to these different forms of identification?”.

Fig 12. Question 19 data.

Percent respondents choosing each percentage interval in response to question 19 “How accurate (in percentage) would this technology need to be in order for you to agree to it being used to identify anyone in society?”.

Fig 13. Questions 20–21 data.

Percent responses to questions 20–21 “Do you think this technology is equally accurate with people of different genders/ethnicities”.

Fig 14. Questions 22–23 data.

Percent respondents agreed (responding 4, 5 or 6 on the scale of agreement) with the use of AFR in these circumstances, if it was more accurate with White than non-White people (questions 22 and 23).