Using AI to improve social housing for the most vulnerable
Using AI to improve social housing for the most vulnerable
Publish Date: 2026-06-23 05:13:00
Source Domain: www.gatescambridge.org
Here’s a summarized list of 6 key points about the article:
-
Project Title and Aims: The “PRISM” (Predictive Risk Intelligence for Social housing Maintenance) project aims to identify vulnerable social housing tenants in UK councils before crises occur.
-
Developing Institutions: The project involves researchers from the University of Cambridge, including Gates Cambridge Scholars Adhib Hussain Syed and Ramit Debnath, and is supported by the Local Government AI Accelerator.
-
Data Integration: PRISM uses three types of data sources for each property: satellite data for heat loss, conventional housing data, and ‘soft’ data such as fuel poverty indicators and tenant contacts.
-
Risk Scoring System: The researchers combine these data to generate a single risk score for each property, identifying both properties in poor condition and the vulnerability of their residents.
-
Impact on Management: The tool is intended to enable more targeted and proactive measures by social housing authorities to prevent deterioration and address related vulnerabilities more efficiently.
-
Human Oversight and Broader Implications: Alerts generated by the model will be reviewed by housing officers rather than acted upon by machines, ensuring human oversight. The project aims to reflect a shift towards better data utilization in social housing regulations and could serve as a model for other councils in the UK.
