2 | Adrian Arnaiz-Rodriguez

Between Help and Harm: An Evaluation of Mental Health Crisis Handling by LLMs

Sun, 01 Mar 2026 00:00:00 +0000

This preprint introduces a unified framework to evaluate how large language models (LLMs) detect and respond to mental health crises, combining a novel taxonomy, a curated dataset, and expert-based assessments. The findings reveal that, while LLMs can offer support in many situations, they also pose significant safety risks — particularly when handling subtle crisis signals. The work highlights critical challenges and provides a foundation for safer, more effective AI-driven mental health interventions.

Studying Causality in Algorithmic Decision Making: the Impact of IA in the Business Domain

Tue, 10 Dec 2024 00:00:00 +0000

Official WebPage of the Public Project: Derechos Digitales en el Trabajo (Digital Rights at Work)

Article written in Spanish for the RRLDE vol.12 n.2: Spanish General Journal of Labor Law and Social Security (Revista General del Derecho del Trabajo y de la Seguridad Social). AI-driven algorithmic management is reshaping workplaces but clashes with existing labor regulations, which are designed for human decision-making. AI relies on correlations, not causes, while labor laws require causal justifications for decisions. This gap can lead to legal insufficiencies and hard-to-detect discrimination. The study explores these issues and suggests possible solutions.

SUMMARY:

Introduction.
Brief history of AI and its evolution.
Use of AI in the business field.
1. Advantages and risks of using AI in decision making.
Concept of AI system. Legal aspects. Technical aspects.
1. Concept of AI system.
2. Legal aspects of AI systems in algorithmic management.
3. Technical aspects of AI systems in algorithmic management.
  1. Operation of AI systems in decision making.
  2. Explanation of the correlation-causation dilemma. Examples in the workplace.
  3. Blackbox. Explainability.
Existence of causality in algorithmic management.
1. Pre-contractual phase.
2. Contractual phase.
Proposal of solutions.
1. 6.1. Technical solutions.
2. Legal solutions.
Conclusions.

Click the Cite button above to show and copy bibtex reference.

The Intersection of Trustworthy AI and Labour Law. A Legal and Technical Study from a Tripartite Taxonomy

Sun, 17 Nov 2024 00:00:00 +0000

Article written in Spanish for the RGDTSS n.69: Spanish General Journal of Labor Law and Social Security (Revista General del Derecho del Trabajo y de la Seguridad Social). This article is a legal and technical study of the intersection between Trustworthy AI and Labour Law. We propose a tripartite taxonomy to understand the implications of the AI Act in the field of Labour Law.

SUMMARY:

Introduction and purpose of the work.
Most relevant legal milestones.
Technical characteristics of sub-symbolic AI.
Life cycle of an AI system: legal obligations.
Ethical principles for a Trustworthy AI: the HLEG ON AI perspective.
Requirements of a Trustworthy AI: intersection with labour law.
Conclusions.

We specifically study the implications of the AI Act and AI technical features in the field of Labour Law in Spain. We just focus on the laboral law.

KEYWORDS: Artificial Intelligence (AI), Labour Law, Labour Law, Reliable AI, AI Regulation, High-Level Expert Group on AI (HLEG), Life Cycle of AI Systems

Click the Cite button above to show and copy bibtex reference.

Mapping the scientific structure of Organization and management of enterprises using complex networks

Mon, 31 Jan 2022 00:00:00 +0000

Click the Cite button above to show and copy bibtex reference.

Videoconferencing Software Options for Telemedicine: A Review for Movement Disorder Neurologists

Mon, 11 Oct 2021 00:00:00 +0000

Click the Cite button above to show and copy bibtex reference.

Analysis of the Learning Process through Eye Tracking Technology and Feature Selection Techniques

Fri, 02 Jul 2021 00:00:00 +0000

Click the Cite button above to show and copy bibtex reference.