Properties of artificial intelligence systems in the context of their use in legal activities
Bibliographic record
Abstract
The study was undertaken to reflect on the values of modern artificial intelligence systems in the context of the use of such systems in legal, especially law enforcement activities.The application of artificial intelligence in the research focuses on the properties of fairness, accountability, and transparency.Fairness should exclude distortions in the operation of artificial intelligence systems, caused by the settings of scales or the specifics of the dataset collected for training the system.Accountability is seen as the property of an AI system to protect user data that is included in a dataset or processed by an AI system.Transparency, on the other hand, reflects the ability to verify the decision logic of an AI system and reverse-engineer its algorithm.This property is currently the least attainable, but it is directly related to the evaluation of the effectiveness of AI, and hence to the possibilities of integrating such systems into legal activities.This paper uses the current understanding of the capabilities of systems based on machine learning methods: convolutional artificial neural networks and transformer networks.The study reveals differences and discussions of AI perspectives in legislation and the state of legal regulation, public and academic approaches to this issue in the European Union, the USA, Canada, Singapore, China, Russia, and Kazakhstan.As a result, the study proposes a set of recommendations for banning/restricting the use of artificial intelligence and decision support systems, considering national and international legislation.
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 | 0.017 | 0.070 |
| Meta-epidemiology (narrow) | 0.000 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.003 |
| Science and technology studies | 0.003 | 0.012 |
| Scholarly communication | 0.011 | 0.017 |
| Open science | 0.001 | 0.004 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.005 | 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".