| --- |
| license: other |
| license_name: ogl-v3 |
| license_link: https://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/ |
| language: |
| - en |
| tags: |
| - uk-government |
| - ai-governance |
| - algorithmic-transparency |
| - responsible-ai |
| - open-data |
| pretty_name: UK Algorithmic Transparency Corpus |
| --- |
| |
| # UK Algorithmic Transparency Corpus |
|
|
| Every published record under the UK's Algorithmic Transparency Recording |
| Standard (ATRS), structured as an open corpus: which public bodies have |
| disclosed which algorithmic and AI tools, and when. |
|
|
| | | | |
| |---|---| |
| | ATRS records | **136** | |
| | Distinct publishing bodies | **73** | |
| | Source | GOV.UK ATRS records | |
| | Licence | Open Government Licence v3.0 | |
|
|
| ## Why |
|
|
| The ATRS is the UK government's standard for publishing how the public sector |
| uses algorithmic tools in decisions that affect people. It is the closest thing |
| to a public register of government AI, and it is growing. But it is published as |
| a set of individual records, not as a dataset you can analyse. This assembles |
| the full set into one structured corpus: publishing body, tool name, description |
| and date per record. |
|
|
| ## The picture |
|
|
| 136 records span **73 distinct public bodies**, led by DSIT, DWP, the Money and |
| Pensions Service, DfE, DESNZ and DBT. The corpus is the base for asking the |
| questions the individual records cannot: which parts of government are most |
| transparent about their AI, what kinds of tools recur, and how disclosure is |
| growing over time. |
|
|
| ## Files |
|
|
| - `data/atrs_records.json` / `.csv` — structured records |
| - `scripts/harvest.py` — reproducible harvest from the GOV.UK Search API |
|
|
| ## Reproduce |
|
|
| ```bash |
| python3 scripts/harvest.py |
| ``` |
|
|
| Contains public sector information licensed under the Open Government Licence |
| v3.0. Independent, self-initiated open research by |
| [Tesseract Academy](https://gov.tesseract.academy). |
|
|
