Design Principles of the AI system

The definition of an AI system it’s in article 3(1) of the AI Act.,

(1) ‘AI system’ means a machine-based system that is designed to operate with varying levels of autonomy and that may exhibit adaptiveness after deployment, and that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments

AI systems are considered high risk systems under the AI Act under three circumstances:

When the AI system is a safety component of a product or, in itself, is a type of product covered by the Union harmonization legislation listed in Annex I (article 6.1(a) of the AI Act)
When the AI system is a safety component of a product or, in itself, is a type of product required to undergo a third party conformity assessment with a view to the placing on the market or the putting into service of that product pursuant Union harmonization legislation listed in Annex I (article 6.1(b) of the AI Act)
When the AI system meets a description of high-risk AI systems included in Annex III of the AI Act (justice systems are listed in point 8 of Annex III).

video_demo_AIJUST_PoC_UPF_16042024_v1c+logo_1920x1080.mp4

The DESIGN PRINCIPLES of the AI judicial assistant for insolvency proceedings for natural persons with no assets are the following:

Automation does not mean uniformity - Task automatization and added value professional human decision
Maximize accuracy through easy detection of errors – providing mechanisms for checking for errors at any given moment of the proceeding
Preserving judicial discretion
Force human action along the way
Minimize data extracted
No identification of individuals

The EXPECTED RESULTS of the AI judicial assistant for insolvency proceedings for natural persons with no assets are the following:

1. Observe whether there’s room for saving time in issuing the final order for forgiving the outstanding debt.
2. Maximize the amount of information based on which the judge adopts its decision and hence minimize the amount of errors, motions and appeals that may be filed.
3. Providing higher transparency and information to third parties as well as society as a whole

The ADDED VALUE of the AI judicial assistant for insolvency proceedings for natural persons with no assets are the following:

1. Experience and assess, through real life experiences such as this proof of concept, the potential and impact of the use of AI and ATM systems in judicial proceedings
2. Test whether the adoption of an AI judicial assistant reduces the time between the claimant files its claim and the issuance of the judicial decision.
3. Observe whether judicial decisions are adopted with more and better information provided by the AI judicial assistant than with human action
4. Observe the implications of document classification and information extraction for better informed judicial decisions based on the potential of large language models
5. Assess the relevance of the compliance requirements set by articles 8 to 15 and 16 to 27 of the AI Act for high-risk systems such as AI systems deployed by judicial systems ([61] of the AI Act and Annex III (8) of the AI Act)
6. Assess the legal implications in terms of obligations, risks and professional liability of considering judges as “deployers of AI systems” under the AI Act.
7. Evaluate the data protection implications of training datasets and ensure their quality and representativeness and determine their representativeness and
8. Address the relevance of the human-machine interaction so that the assistant supports and improves, but does not replace judicial decisions adopted by humans
9. Be aware of the risks and potential errors these types of systems have and learn how to detect them.
10. Address the possible biases these systems may have, detect them and correct them.
11. Potential to bring higher transparency to credit markets and to public registries
12. Capacity to observe, in real time, the evolution and type of cases so that policy and courts can react in a faster way.
13. Respect linguistic diversity and issue court documents and judicial decisions in multiple languages.

Page updated

Google Sites

Report abuse