Statistics and Society

Statistics and Society

Date: Wednesday 01 December 2021, 3.00PM
Location: Online
Special Interests Group Meeting

Share this event

Algorithms shape our everyday lives and impact policy and society. The realisation that not all of these effects are beneficial is well established in both machine learning and statistics. This meeting explores this set-up with a talk by a technology-aware lawyer (Marion Oswald), and one of the original contributors on the discussion of algorithmic fairness (Alexandra Chouldechova). 
3.00 pm Good law, good science and accountability: Achieving trustworthy use of emerging technologies in UK policing - Dr Marion Oswald
3.45 pm - 4 pm Break
4 pm-4.45 pm Dr Alexandra Chouldechova TBA

Good law, good science and accountability: Achieving trustworthy use of emerging technologies in UK policing - Dr Marion Oswald

There is no shortage of ethical standards for AI and data analytics but most have yet to be translated into operational guidance (Babuta, Oswald and Janjeva, 2020).  Frameworks tend to focus upon project methodology in an attempt to make it ‘more ethical’, potentially obviating personal responsibility
for assessing the validity and wider consequences of AI objectively using all relevant evidence and factors.  But no amount of ethical discussion can alleviate the inconvenient truth that policing technology innovation must be within the boundaries of the law and good science before further consideration of ethics is valuable.

The policing technology landscape has been described as ‘very confused’ with ‘no single regulatory voice’ and much uncertainty as to whether the police have ‘sufficient public trust to take things forward on their own without being accused of marking their own homework’. Themes emerging from proceedings of the West Midlands Police data ethics committee (which Oswald chairs) demonstrate that the technical and statistical aspects of data analytics should not be isolated from the legal, contextual, operational and ethical considerations, as each will influence the other. 
This paper proposes a three-pillar approach to achieving trustworthy use of emerging technologies in the UK policing: first, governing law plus guidance/policy interpreted for the relevant context; secondly, standards, both ethical standards attached to personal responsibility and scientific standards; and thirdly, the need for people at all levels within the policing body who are committed to accountability; all of which should be subject to rolling independent oversight.
RSS Women in Data Science and Statistics special interest group
Sofia Olhede