Abstract
Summary:
1. We strongly support the undertaking of a thematic review on science and technology, including the emerging use of technologies and in particular, those that use Artificial Intelligence, and those that provide probabilistic outputs. This should be combined with the theme of data analytics.
2. We strongly support the introduction of specific inspection of the use of science and technological solutions in policing as part of PEEL assessments to enhance efforts to ensure the upholding of the NPCC AI in Policing Covenant and the adoption of responsible technologies. In order to ensure that inspections cover police efficiency, effectiveness and legitimacy, technical and statistical aspects of policing should not be separated from legal, contextual, operational, and ethical considerations. As laid out in the new checklist for Responsible AI in Policing (appended), relevant factors to judge efficiency, effectiveness and legitimacy regarding the use of advanced data analytics and AI should cover i) technical validity, ii) operational deployment, and iii) legality and proportionality. The factors in this checklist could provide a template for questions to be asked in any new inspection regime focused on AI and data analytics.
1. We strongly support the undertaking of a thematic review on science and technology, including the emerging use of technologies and in particular, those that use Artificial Intelligence, and those that provide probabilistic outputs. This should be combined with the theme of data analytics.
2. We strongly support the introduction of specific inspection of the use of science and technological solutions in policing as part of PEEL assessments to enhance efforts to ensure the upholding of the NPCC AI in Policing Covenant and the adoption of responsible technologies. In order to ensure that inspections cover police efficiency, effectiveness and legitimacy, technical and statistical aspects of policing should not be separated from legal, contextual, operational, and ethical considerations. As laid out in the new checklist for Responsible AI in Policing (appended), relevant factors to judge efficiency, effectiveness and legitimacy regarding the use of advanced data analytics and AI should cover i) technical validity, ii) operational deployment, and iii) legality and proportionality. The factors in this checklist could provide a template for questions to be asked in any new inspection regime focused on AI and data analytics.
| Original language | English |
|---|---|
| Publisher | HMICFRS |
| Number of pages | 3 |
| Place of Publication | Birmingham |
| Publication status | Accepted/In press - 25 Oct 2024 |