Government by Algorithms at the Light of Freedom of Information Regimes: A Case-by-Case Approach on ADM Systems within Public Education Sector
Bibliographic record
Abstract
Abstract: What the Houston Court qualified as "mysterious 'black box' impervious to challenge" was in practice a sophisticated software of many layers of calculations, which rated teachers' effectiveness to make employment decisions. In the European Union, a system as such would fall under the Proposal for AI Regulation of 2021, which qualifies AI models in education and vocational training as "high-risk" systems. Automated decision-making systems (ADM systems), AI-driven or not, are being increasingly used by governments in public education for different purposes, such as handling applications for undergraduate admission or profiling students and teachers to assess their performance. Across cases and jurisdictions, there is growing evidence of how the use of ADM systems in the education sector is becoming quite problematic: arbitrary assignment of teaching posts in mobility procedures, undue barriers to access undergraduate studies, and frequent lack of transparency in their implementation and decisions. This Article discusses how Freedom of Information Act (FOIA) regimes may contribute to rendering governments' ADM systems (AI-driven or not) accountable. The analysis of the FOIA cases (Parcoursoup saga in France, MIUR in Italy, and Ofqual in the United Kingdom) shows to what extent decisions granting access to the source code, functional and technical specifications, or third-party audits allow public scrutiny of ADM systems, detection of their pathologies, and better understanding of their adverse impacts on rights and freedoms, individual or collective. This Article also addresses the constitutional value of the right of access to public records (Parcoursup), and the importance of proactive and mandatory public dissemination to ensure traceability, transparency, and accountability of the ADM systems for FOIA purposes. In this sense, some legal initiatives across jurisdictions (Canada, France, Spain, United States, European Union) enhancing transparency and accountability of algorithmic systems will be examined.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.043 | 0.058 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.005 | 0.004 |
| Science and technology studies | 0.008 | 0.027 |
| Scholarly communication | 0.019 | 0.015 |
| Open science | 0.003 | 0.009 |
| Research integrity | 0.012 | 0.009 |
| Insufficient payload (model declined to judge) | 0.010 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".