Transparency mechanisms for UK public-sector algorithmic decision-making systems

We currently have a fragmented landscape of mechanisms for transparency that, taken
individually or combined in the limited ways currently possible, leave us far from ensuring that
we are capable of scrutinising and evaluating the functions, or effects on communities and
individuals, of ADM systems in use or under consideration in central or local government.

In order to obtain meaningful transparency there are some key questions that need to be answered, for example:

• What are the data sources of a system?
• Who is operating it?
• What are the conditions around repurposing or using it in other contexts?
• Why is the system being developed?
• What were the alternatives and how was a specific ADM system selected?
• What is the logic of the system?
• Who built it and how much did it cost?
• Were private actors involved?
• Are there specific impacted groups, and how are they impacted?

This information is nominally available in transparency documents that are produced by local and central government and publicsector officials. In practice these documents and information are seldom released into the public domain, and if they are released, are not organised or made available in ways that make it easy to consult.

These limitations are compounded by the reality that some existing transparency protocols are not currently mandatory, or not mandatory for ADM systems; some are mandatory but not effectively enforced; and those that are mandatory could be better promoted through best-practice guidelines that offer practical support to overresourced public authorities.

To create coherence out of this landscape, and systematically strengthen transparency practices across government, we have focused on currently underused informational assets that, taken together (if not individually), can enable us to make meaningful extractions and inferences about the way ADM systems are used across government.