The Centre for Data Ethics and Innovation (CDEI) published this blog in early July 2021 about workshops it had run recently with the public who considered what algorithmic transparency should look like - with a view to developing a 'transparency obligation' for public sector organisations. In the third phase of the research, the public designed a prototype information format, assigning 'categories' of information into two tiers. The first tier is where algorithms are high risk or impact, and the second covers all algorithms. It is likely of interest with its categories of information, which include things like: purpose, contact, risks, commercial information, etc.