Police use Experian Marketing Data for AI Custody Decisions

Durham Police has paid global data broker Experian for UK postcode stereotypes built on 850 million pieces of information to feed into an artificial intelligence (AI) tool used in custody decisions, a Big Brother Watch investigation has revealed.

 

Durham Police is feeding Experian’s ‘Mosaic’ data, which profiles all 50 million adults in the UK[1] to classify UK postcodes, households and even individuals[2] into stereotypes, into its AI ‘Harm Assessment Risk Tool’ (HART). The 66 ‘Mosaic’ categories include ‘Disconnected Youth’, ‘Asian Heritage’ and ‘Dependent Greys’.[3]

Durham Police’s AI tool processes Experian’s ‘Mosaic’ data and other personal information to predict whether a suspect might be at low, medium or high risk of reoffending.[4]

Experian’s Mosaic code includes the ‘demographic characteristics’ of each stereotype – characterising ‘Asian Heritage’ as ‘extended families’ living in ‘inexpensive, close-packed Victorian terraces’, adding that ‘when people do have jobs, they are generally in low paid routine occupations in transport or food service’.[5]

‘Disconnected Youth’ are characterised as ‘avid texters’ whose ‘wages are often low’[6] – with first names like ‘Liam’ and ‘Chelsea’.[7]

Durham Police’s AI risk predictions guide decisions as to whether a suspect should be charged or released onto the ‘Checkpoint’ rehabilitation programme.[8] [9] Moderate risk’ suspects[10] are informed that if they successfully complete the Checkpoint programme they will not receive a criminal conviction.[11]

According to an Experian presentation obtained by Big Brother Watch, ‘Mosaic’ data profiles ‘all 50 million adults in the UK’ based on over ‘850 million pieces of data’ including:

 

• Family composition, including children,
• Family/personal names linked to ethnicity,
• Online data, including data scraped from the pregnancy advice website ‘Emma’s Diary’, and Rightmove,
• Occupation,
• Child benefits, tax credits, and income support,
• Health data,
• GCSE results,
• Ratio of gardens to buildings,
• Census data,
• Gas and electricity consumption. [12]

 

A register of contracts obtained by Big Brother Watch reveals that Durham Police paid £45,913 to Experian, including £25,913 for the ‘Mosaic’ system.[13]

Experian’s ‘Mosaic’ links names to stereotypes: for example, people called ‘Stacey’ are likely to fall under ‘Families with Needs’ who receive ‘a range of benefits’; ‘Abdi’ and ‘Asha’ are ‘Crowded Kaleidoscope’ described as ‘multi-cultural’ families likely to live in ‘cramped’ and ‘overcrowded flats’; whilst ‘Terrence’ and ‘Denise’ are ‘Low-Income Workers’ who have ‘few qualifications’ and are ‘heavy TV viewers’.[14]

Silkie Carlo, Director of Big Brother Watch, said:

“For a credit checking company to collect millions of pieces of information about us and sell profiles to the highest bidder is chilling. But for police to feed these crude and offensive profiles through artificial intelligence to make decisions on freedom and justice in the UK is truly dystopian.

“We wouldn’t accept people going through our bins to collect information about us. Nor should we accept multi-billion pound companies like Experian scavenging for information about us online or offline, whether for profit or policing.

Parliament should urgently consider what place this big data and artificial intelligence has in our policing.”

 

Sheena Urwin, Head of Criminal Justice at Durham Constabulary, said:

“The force entered into a contract with Experian using Mosaic Public Sector to better understand our communities and to improve our engagement – the data they provided helped us do that. Our aim is to reduce harm to the communities we serve and improve life chances for the people we come into contact with.

“We began a research project with Cambridge University around the Harm Assessment Risk Tool (HART) to try to understand whether we could predict an offender’s risk of reoffending, in order to provide appropriate support to encourage them away from a life of crime. One of the variables used in the initial model was the Mosaic code.

“We must stress that the HART tool contributes to the decision-making process by assessing the risk of reoffending; however, the final decision remains with the custody sergeant who factors in the statutory considerations outlined under both the Policing and Crime Act 2017 and the Bail Act 1976. We are continuing to evaluate the research with our academic partners.”

 

References:

Experian’s Mosaic Public Sector brochure: http://www.experian.co.uk/assets/marketing-services/brochures/mosaic-ps-brochure.pdf

[1] Mosaic Infographic, Experian, http://www.experian.co.uk/marketing-services/knowledge/infographics/infographic-new-mosaic.html  (accessed 28 March 2018)

[2] Ibid. Also see Under the bonnet: Mosaic data, methodology and build, Paul Cresswell et al., Experian Marketing Services, 1 April 2014, p.7: http://www.experian.co.uk/assets/marketing-services/presentations/mosaic-data-methodology-and-build.pdf

[3] Mosaic Public Sector brochure, Experian, 2016, pp.6-9: http://www.experian.co.uk/assets/marketing-services/brochures/mosaic-ps-brochure.pdf

[4] ALGORITHMIC FORECASTING OF OFFENDER DANGEROUSNESS FOR POLICE CUSTODY OFFICERS: AN ASSESSMENT OF ACCURACY FOR THE DURHAM CONSTABULARY MODEL – Sheena Urwin, December 2016, p.98: http://www.crim.cam.ac.uk/alumni/theses/Sheena%20Urwin%20Thesis%2012-12-2016.pdf

[5] Mosaic UK Data Profile, Experian, 2017, p.51: https://www.experianintact.com/content/uk/documents/productSheets/MosaicConsumerUK.pdf

[6] Mosaic UK Data Profile, Experian, 2017, p.41: https://www.experianintact.com/content/uk/documents/productSheets/MosaicConsumerUK.pdf

[7] Mosaic UK Data Profile, Experian, 2017, p.6: https://www.experianintact.com/content/uk/documents/productSheets/MosaicConsumerUK.pdf

[8] Helping police make custody decisions using artificial intelligence, University of Cambridge, 26 Feb 2018: http://www.cam.ac.uk/research/features/helping-police-make-custody-decisions-using-artificial-intelligence

[9] Checkpoint, Durham Constabulary (accessed 29 March 2018): https://www.durham.police.uk/Information-and-advice/Pages/Checkpoint.aspx

[10] Algorithmic risk assessment policing models: Lessons from the Durham Constabulary HART model, M. Oswald, J. Grace, S. Urwin (Durham Constabulary) & G.C. Barnes, 31 August 2017, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3029345

[11] Script for Custody Officers (Checkpoint), Durham Police, https://www.scribd.com/document/372200432/Script-for-Custody-Officers-Durham-Police

[12] Under the bonnet: Mosaic data, methodology and build, Paul Cresswell et al., Experian Marketing Services, 1 April 2014: http://www.experian.co.uk/assets/marketing-services/presentations/mosaic-data-methodology-and-build.pdf

[13] Durham PCC Register of Contracts (accessed 29 March 2018): https://www.durham-pcc.gov.uk/document-library/finance/register-of-contractspcc.pdf

[14] Mosaic UK Data Profile, Experian, 2017: https://www.experianintact.com/content/uk/documents/productSheets/MosaicConsumerUK.pdf

 

 

---