analiza aiqbench
This research extends algorithmic harm taxonomies to address novel causal drivers and endogenous risks like sanitization when generative models are integrated into recommendation systems. By analyzing problematic subsets of the input-output joint distribution, the study provides a framework for detecting and mitigating these specific harms in non-conversational ML platforms. The work is essential for advancing safety protocols as generative AI becomes more deeply embedded in core machine learning infrastructure.
Algorithmic Harms Associated with Generative Model-Augmented Recommendation Systems
arXiv:2609.33073v1 Announce Type: new Abstract: In this work, we consider algorithmic harms that may arise as generative models are incorporated into machine learning platforms. We argue that existing harm taxonomies and threat models require extension to (1) address novel causal drivers of well-studied representational and quality-of-service harms; and (2) anticipate and mitigate endogenous harms, such as sanitization, which may arise when system inputs are misaligned with the system designer's objectives, or the generative model's inductive priors. To this end, we introduce an expanded taxonomy of algorithmic harms associated with the use of generative models in non-conversational recommendation systems. In addition, we offer a causal analysis of how problematic subsets of the (input, output) joint distribution can arise, in an effort to inform harms detection and mitigation efforts.
Abstract:In this work, we consider algorithmic harms that may arise as generative models are incorporated into machine learning platforms. We argue that existing harm taxonomies and threat models require extension to (1) address novel causal drivers of well-studied representational and quality-of-service harms; and (2) anticipate and mitigate endogenous harms, such as sanitization, which may arise when system inputs are misaligned with the system designer's objectives, or the generative model's inductive priors. To this end, we introduce an expanded taxonomy of algorithmic harms associated with the use of generative models in non-conversational recommendation systems. In addition, we offer a causal analysis of how problematic subsets of the (input, output) joint distribution can arise, in an effort to inform harms detection and mitigation efforts.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Information Retrieval (cs.IR)
arXiv:2609.33073 [cs.LG]
(or arXiv:2609.33073v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2609.33073 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 226 cuvinte
Subiecte conexe
Berkeley Law report finds AI use affects rights in law, media, schools
Quantum Zeitgeist · 29 sept. 2026
The J-Space Debate, Agent Swarms, and Pacing Frontier AI - Digital Minds Newsletter #4
LessWrong · 18 sept. 2026
Crop Recommendation and Agricultural Query Answering System Using Spatio-Temporal Graph Neural Networks and Hybrid Retrieval Augmentation
arXiv · 9 iun. 2026
The world's best gradual disempowerment model organism: Frontier AI labs
LessWrong · 29 sept. 2026