University of Padua
Abstract
Large Language Models (LLMs) can generate highly persuasive text, raising concerns about their misuse for propaganda, manipulation, and other harmful purposes. This leads us to our central question: Is LLM-generated persuasion more difficult to automatically detect than human-written persuasion? To address this, we categorize controllable generation approaches for producing persuasive content with LLMs and introduce Persuaficial, a high-quality multilingual benchmark covering six languages: English, German, Polish, Italian, French and Russian. Using this benchmark, we conduct extensive empirical evaluations comparing human-authored and LLM-generated persuasive texts. We find that although overtly persuasive LLM-generated texts can be easier to detect than human-written ones, subtle LLM-generated persuasion consistently degrades automatic detection performance. Beyond detection performance, we provide the first comprehensive linguistic analysis contrasting human and LLM-generated persuasive texts, offering insights that may guide the development of more interpretable and robust detection tools.
Large Language Models (LLMs) can generate highly persuasive text, raising concerns about their misuse for propaganda, manipulation, and other harmful purposes. This leads us to our central question: Is LLM-generated persuasion more difficult to automatically detect than human-written persuasion? To address this, we categorize controllable generation approaches for producing persuasive content with LLMs and introduce Persuaficial, a high-quality multilingual benchmark covering six languages: English, German, Polish, Italian, French and Russian. Using this benchmark, we conduct extensive empirical evaluations comparing human-authored and LLM-generated persuasive texts. We find that although overtly persuasive LLM-generated texts can be easier to detect than human-written ones, subtle LLM-generated persuasion consistently degrades automatic detection performance. Beyond detection performance, we provide the first comprehensive linguistic analysis contrasting human and LLM-generated persuasive texts, offering insights that may guide the development of more interpretable and robust detection tools.
Why we are recommending this paper?
Due to your Interest in AI Bias
This paper directly addresses concerns about AI manipulation and persuasion, aligning with the user’s interest in AI bias and data transparency. It investigates a critical question regarding the detectability of LLM-generated text, a key area for mitigating potential harms.
Due to your Interest in AI Bias
This paper directly addresses concerns about AI manipulation and persuasion, aligning with the user’s interest in AI bias and data transparency. It investigates a critical question regarding the detectability of LLM-generated text, a key area for mitigating potential harms.
University of Cambridge
Abstract
The era of AI regulation (AIR) is upon us. But AI systems, we argue, will not be able to comply with these regulations at the necessary speed and scale by continuing to rely on traditional, analogue methods of compliance. Instead, we posit that compliance with these regulations will only realistically be achieved computationally: that is, with algorithms that run across the life cycle of an AI system, automatically steering it toward AIR compliance in the face of dynamic conditions. Yet despite their (we would argue) inevitability, the research community has yet to specify exactly how these algorithms for computational AIR compliance should behave - or how we should benchmark their performance. To fill these gaps, we specify a set of design goals for such algorithms. In addition, we specify a benchmark dataset that can be used to quantitatively measure whether individual algorithms satisfy these design goals. By delivering this blueprint, we hope to give shape to an important but uncrystallized new domain of research - and, in doing so, incite necessary investment in it.
The era of AI regulation (AIR) is upon us. But AI systems, we argue, will not be able to comply with these regulations at the necessary speed and scale by continuing to rely on traditional, analogue methods of compliance. Instead, we posit that compliance with these regulations will only realistically be achieved computationally: that is, with algorithms that run across the life cycle of an AI system, automatically steering it toward AIR compliance in the face of dynamic conditions. Yet despite their (we would argue) inevitability, the research community has yet to specify exactly how these algorithms for computational AIR compliance should behave - or how we should benchmark their performance. To fill these gaps, we specify a set of design goals for such algorithms. In addition, we specify a benchmark dataset that can be used to quantitatively measure whether individual algorithms satisfy these design goals. By delivering this blueprint, we hope to give shape to an important but uncrystallized new domain of research - and, in doing so, incite necessary investment in it.
Why we are recommending this paper?
Due to your Interest in AI Transparency
Given the user's interest in AI ethics and transparency, this paper’s exploration of agentic AI systems – incorporating planning and reasoning – is highly relevant. It’s a timely discussion about the evolving capabilities of AI and their potential ethical implications.
Due to your Interest in AI Transparency
Given the user's interest in AI ethics and transparency, this paper’s exploration of agentic AI systems – incorporating planning and reasoning – is highly relevant. It’s a timely discussion about the evolving capabilities of AI and their potential ethical implications.
The Hong Kong Polytechnic University
Abstract
Although machine unlearning is essential for removing private, harmful, or copyrighted content from LLMs, current benchmarks often fail to faithfully represent the true "forgetting scope" learned by the model. We formalize two distinct unlearning granularities, domain-level and instance-level, and propose BiForget, an automated framework for synthesizing high-quality forget sets. Unlike prior work relying on external generators, BiForget exploits the target model per se to elicit data that matches its internal knowledge distribution through seed-guided and adversarial prompting. Our experiments across diverse benchmarks show that it achieves a superior balance of relevance, diversity, and efficiency. Quantitatively, in the Harry Potter domain, it improves relevance by ${\sim}20$ and diversity by ${\sim}$0.05 while halving the total data size compared to SOTAs. Ultimately, it facilitates more robust forgetting and better utility preservation, providing a more rigorous foundation for evaluating LLM unlearning.
Although machine unlearning is essential for removing private, harmful, or copyrighted content from LLMs, current benchmarks often fail to faithfully represent the true "forgetting scope" learned by the model. We formalize two distinct unlearning granularities, domain-level and instance-level, and propose BiForget, an automated framework for synthesizing high-quality forget sets. Unlike prior work relying on external generators, BiForget exploits the target model per se to elicit data that matches its internal knowledge distribution through seed-guided and adversarial prompting. Our experiments across diverse benchmarks show that it achieves a superior balance of relevance, diversity, and efficiency. Quantitatively, in the Harry Potter domain, it improves relevance by ${\sim}20$ and diversity by ${\sim}$0.05 while halving the total data size compared to SOTAs. Ultimately, it facilitates more robust forgetting and better utility preservation, providing a more rigorous foundation for evaluating LLM unlearning.
Why we are recommending this paper?
Due to your Interest in Data Representation
This research tackles the critical issue of data bias within LLMs, directly relating to the user’s interest in data fairness and transparency. The concept of granular unlearning aligns with the need to understand and mitigate hidden biases in training data.
Due to your Interest in Data Representation
This research tackles the critical issue of data bias within LLMs, directly relating to the user’s interest in data fairness and transparency. The concept of granular unlearning aligns with the need to understand and mitigate hidden biases in training data.
Fraunhofer Portugal AICOS
Abstract
Evaluating machine learning (ML) model bias is key to building trustworthy and robust ML systems. Counterfactual Fairness (CF) audits allow the measurement of bias of ML models with a causal framework, yet their conclusions rely on a single causal graph that is rarely known with certainty in real-world scenarios. We propose CF with Graph Uncertainty (CF-GU), a bias evaluation procedure that incorporates the uncertainty of specifying a causal graph into CF. CF-GU (i) bootstraps a Causal Discovery algorithm under domain knowledge constraints to produce a bag of plausible Directed Acyclic Graphs (DAGs), (ii) quantifies graph uncertainty with the normalized Shannon entropy, and (iii) provides confidence bounds on CF metrics. Experiments on synthetic data show how contrasting domain knowledge assumptions support or refute audits of CF, while experiments on real-world data (COMPAS and Adult datasets) pinpoint well-known biases with high confidence, even when supplied with minimal domain knowledge constraints.
Evaluating machine learning (ML) model bias is key to building trustworthy and robust ML systems. Counterfactual Fairness (CF) audits allow the measurement of bias of ML models with a causal framework, yet their conclusions rely on a single causal graph that is rarely known with certainty in real-world scenarios. We propose CF with Graph Uncertainty (CF-GU), a bias evaluation procedure that incorporates the uncertainty of specifying a causal graph into CF. CF-GU (i) bootstraps a Causal Discovery algorithm under domain knowledge constraints to produce a bag of plausible Directed Acyclic Graphs (DAGs), (ii) quantifies graph uncertainty with the normalized Shannon entropy, and (iii) provides confidence bounds on CF metrics. Experiments on synthetic data show how contrasting domain knowledge assumptions support or refute audits of CF, while experiments on real-world data (COMPAS and Adult datasets) pinpoint well-known biases with high confidence, even when supplied with minimal domain knowledge constraints.
Why we are recommending this paper?
Due to your Interest in Data Fairness
The paper’s focus on counterfactual fairness audits and causal graphs directly addresses the user’s interest in AI fairness and bias detection. Understanding how causal relationships are used in bias evaluation is crucial for building trustworthy AI systems.
Due to your Interest in Data Fairness
The paper’s focus on counterfactual fairness audits and causal graphs directly addresses the user’s interest in AI fairness and bias detection. Understanding how causal relationships are used in bias evaluation is crucial for building trustworthy AI systems.
None
Abstract
In this report, we investigate the potential use of large language models (LLM's) in the task of data compression. Previous works have demonstrated promising results in applying LLM's towards compressing not only text, but also a wide range of multi-modal data. Despite the favorable performance achieved, there still remains several practical questions that pose a challenge towards replacing existing data compression algorithms with LLM's. In this work, we explore different methods to achieve a lower adjusted compression rate using LLM's as data compressors. In comparison to previous works, we were able to achieve a new state-of-the-art (SOTA) adjusted compression rate of around $18\%$ on the enwik9 dataset without additional model training. Furthermore, we explore the use of LLM's in compressing non-English data, code data, byte stream sequences. We show that while LLM's excel in compressing data in text-dominant domains, their ability in compressing non-natural text sequences still remain competitive if configured in the right way.
In this report, we investigate the potential use of large language models (LLM's) in the task of data compression. Previous works have demonstrated promising results in applying LLM's towards compressing not only text, but also a wide range of multi-modal data. Despite the favorable performance achieved, there still remains several practical questions that pose a challenge towards replacing existing data compression algorithms with LLM's. In this work, we explore different methods to achieve a lower adjusted compression rate using LLM's as data compressors. In comparison to previous works, we were able to achieve a new state-of-the-art (SOTA) adjusted compression rate of around $18\%$ on the enwik9 dataset without additional model training. Furthermore, we explore the use of LLM's in compressing non-English data, code data, byte stream sequences. We show that while LLM's excel in compressing data in text-dominant domains, their ability in compressing non-natural text sequences still remain competitive if configured in the right way.
Why we are recommending this paper?
Due to your Interest in Data Representation
This paper explores the potential of LLMs for data compression, a technique that can impact data representation and transparency. Given the user's interest in data representation and fairness, understanding how LLMs can be used to manage and control data is valuable.
Due to your Interest in Data Representation
This paper explores the potential of LLMs for data compression, a technique that can impact data representation and transparency. Given the user's interest in data representation and fairness, understanding how LLMs can be used to manage and control data is valuable.
Karlsruhe Institute of Technology
Abstract
Sampling is renowned for its privacy amplification in differential privacy (DP), and is often assumed to improve the utility of a DP mechanism by allowing a noise reduction. In this paper, we further show that this last assumption is flawed: When measuring utility at equal privacy levels, sampling as preprocessing consistently yields penalties due to utility loss from omitting records over all canonical DP mechanisms -- Laplace, Gaussian, exponential, and report noisy max -- as well as recent applications of sampling, such as clustering. Extending this analysis, we investigate suppression as a generalized method of choosing, or omitting, records. Developing a theoretical analysis of this technique, we derive privacy bounds for arbitrary suppression strategies under unbounded approximate DP. We find that our tested suppression strategy also fails to improve the privacy-utility tradeoff. Surprisingly, uniform sampling emerges as one of the best suppression methods -- despite its still degrading effect. Our results call into question common preprocessing assumptions in DP practice.
Sampling is renowned for its privacy amplification in differential privacy (DP), and is often assumed to improve the utility of a DP mechanism by allowing a noise reduction. In this paper, we further show that this last assumption is flawed: When measuring utility at equal privacy levels, sampling as preprocessing consistently yields penalties due to utility loss from omitting records over all canonical DP mechanisms -- Laplace, Gaussian, exponential, and report noisy max -- as well as recent applications of sampling, such as clustering. Extending this analysis, we investigate suppression as a generalized method of choosing, or omitting, records. Developing a theoretical analysis of this technique, we derive privacy bounds for arbitrary suppression strategies under unbounded approximate DP. We find that our tested suppression strategy also fails to improve the privacy-utility tradeoff. Surprisingly, uniform sampling emerges as one of the best suppression methods -- despite its still degrading effect. Our results call into question common preprocessing assumptions in DP practice.
Why we are recommending this paper?
Due to your Interest in Data Ethics
Due to your Interest in Data Ethics
Hebrew University
Abstract
Alignment of artificial intelligence (AI) encompasses the normative problem of specifying how AI systems should act and the technical problem of ensuring AI systems comply with those specifications. To date, AI alignment has generally overlooked an important source of knowledge and practice for grappling with these problems: law. In this paper, we aim to fill this gap by exploring how legal rules, principles, and methods can be leveraged to address problems of alignment and inform the design of AI systems that operate safely and ethically. This emerging field -- legal alignment -- focuses on three research directions: (1) designing AI systems to comply with the content of legal rules developed through legitimate institutions and processes, (2) adapting methods from legal interpretation to guide how AI systems reason and make decisions, and (3) harnessing legal concepts as a structural blueprint for confronting challenges of reliability, trust, and cooperation in AI systems. These research directions present new conceptual, empirical, and institutional questions, which include examining the specific set of laws that particular AI systems should follow, creating evaluations to assess their legal compliance in real-world settings, and developing governance frameworks to support the implementation of legal alignment in practice. Tackling these questions requires expertise across law, computer science, and other disciplines, offering these communities the opportunity to collaborate in designing AI for the better.
Alignment of artificial intelligence (AI) encompasses the normative problem of specifying how AI systems should act and the technical problem of ensuring AI systems comply with those specifications. To date, AI alignment has generally overlooked an important source of knowledge and practice for grappling with these problems: law. In this paper, we aim to fill this gap by exploring how legal rules, principles, and methods can be leveraged to address problems of alignment and inform the design of AI systems that operate safely and ethically. This emerging field -- legal alignment -- focuses on three research directions: (1) designing AI systems to comply with the content of legal rules developed through legitimate institutions and processes, (2) adapting methods from legal interpretation to guide how AI systems reason and make decisions, and (3) harnessing legal concepts as a structural blueprint for confronting challenges of reliability, trust, and cooperation in AI systems. These research directions present new conceptual, empirical, and institutional questions, which include examining the specific set of laws that particular AI systems should follow, creating evaluations to assess their legal compliance in real-world settings, and developing governance frameworks to support the implementation of legal alignment in practice. Tackling these questions requires expertise across law, computer science, and other disciplines, offering these communities the opportunity to collaborate in designing AI for the better.
Why we are recommending this paper?
Due to your Interest in Data Ethics
Due to your Interest in Data Ethics
Scuola Superiore SantAnna
Abstract
Group-based reinforcement learning methods, like Group Relative Policy Optimization (GRPO), are widely used nowadays to post-train large language models. Despite their empirical success, they exhibit structural mismatches between reward optimization and the underlying training objective. In this paper, we present a theoretical analysis of GRPO style methods by studying them within a unified surrogate formulation. This perspective reveals recurring properties that affect all the methods under analysis: (i) non-uniform group weighting induces systematic gradient biases on shared prefix tokens; (ii) interactions with the AdamW optimizer make training dynamics largely insensitive to reward scaling; and (iii) optimizer momentum can push policy updates beyond the intended clipping region under repeated optimization steps. We believe that these findings highlight fundamental limitations of current approaches and provide principled guidance for the design of future formulations.
Group-based reinforcement learning methods, like Group Relative Policy Optimization (GRPO), are widely used nowadays to post-train large language models. Despite their empirical success, they exhibit structural mismatches between reward optimization and the underlying training objective. In this paper, we present a theoretical analysis of GRPO style methods by studying them within a unified surrogate formulation. This perspective reveals recurring properties that affect all the methods under analysis: (i) non-uniform group weighting induces systematic gradient biases on shared prefix tokens; (ii) interactions with the AdamW optimizer make training dynamics largely insensitive to reward scaling; and (iii) optimizer momentum can push policy updates beyond the intended clipping region under repeated optimization steps. We believe that these findings highlight fundamental limitations of current approaches and provide principled guidance for the design of future formulations.
Why we are recommending this paper?
Due to your Interest in AI Bias
Due to your Interest in AI Bias
Tokyo Institute of Technology
Abstract
Training data is a critical and often proprietary asset in Large Language Model (LLM) development, motivating the use of data watermarking to embed model-transferable signals for usage verification. We identify low coverage as a vital yet largely overlooked requirement for practicality, as individual data owners typically contribute only a minute fraction of massive training corpora. Prior methods fail to maintain stealthiness, verification feasibility, or robustness when only one or a few sequences can be modified. To address these limitations, we introduce SLIM, a framework enabling per-user data provenance verification under strict black-box access. SLIM leverages intrinsic LLM properties to induce a Latent-Space Confusion Zone by training the model to map semantically similar prefixes to divergent continuations. This manifests as localized generation instability, which can be reliably detected via hypothesis testing. Experiments demonstrate that SLIM achieves ultra-low coverage capability, strong black-box verification performance, and great scalability while preserving both stealthiness and model utility, offering a robust solution for protecting training data in modern LLM pipelines.
Training data is a critical and often proprietary asset in Large Language Model (LLM) development, motivating the use of data watermarking to embed model-transferable signals for usage verification. We identify low coverage as a vital yet largely overlooked requirement for practicality, as individual data owners typically contribute only a minute fraction of massive training corpora. Prior methods fail to maintain stealthiness, verification feasibility, or robustness when only one or a few sequences can be modified. To address these limitations, we introduce SLIM, a framework enabling per-user data provenance verification under strict black-box access. SLIM leverages intrinsic LLM properties to induce a Latent-Space Confusion Zone by training the model to map semantically similar prefixes to divergent continuations. This manifests as localized generation instability, which can be reliably detected via hypothesis testing. Experiments demonstrate that SLIM achieves ultra-low coverage capability, strong black-box verification performance, and great scalability while preserving both stealthiness and model utility, offering a robust solution for protecting training data in modern LLM pipelines.
Why we are recommending this paper?
Due to your Interest in Data Transparency
Due to your Interest in Data Transparency
Prince Sultan University
AI Insights
- The paper discusses the concept of agentic AI, which refers to AI systems that can make decisions and take actions on their own without human intervention. [3]
- Autonomous agent: An AI system that can operate independently and make decisions based on its internal state and external environment. [3]
- Agentic AI is a rapidly growing field with various applications, including scientific discovery, language translation, and task management. [2]
Abstract
The evolution of Large Language Models (LLMs) from passive text generators to autonomous, goal-driven systems represents a fundamental shift in artificial intelligence. This chapter examines the emergence of agentic AI systems that integrate planning, memory, tool use, and iterative reasoning to operate autonomously in complex environments. We trace the architectural progression from statistical models to transformer-based systems, identifying capabilities that enable agentic behavior: long-range reasoning, contextual awareness, and adaptive decision-making. The chapter provides three contributions: (1) a synthesis of how LLM capabilities extend toward agency through reasoning-action-reflection loops; (2) an integrative framework describing core components perception, memory, planning, and tool execution that bridge LLMs with autonomous behavior; (3) a critical assessment of applications and persistent challenges in safety, alignment, reliability, and sustainability. Unlike existing surveys, we focus on the architectural transition from language understanding to autonomous action, emphasizing the technical gaps that must be resolved before deployment. We identify critical research priorities, including verifiable planning, scalable multi-agent coordination, persistent memory architectures, and governance frameworks. Responsible advancement requires simultaneous progress in technical robustness, interpretability, and ethical safeguards to realize potential while mitigating risks of misalignment and unintended consequences.
The evolution of Large Language Models (LLMs) from passive text generators to autonomous, goal-driven systems represents a fundamental shift in artificial intelligence. This chapter examines the emergence of agentic AI systems that integrate planning, memory, tool use, and iterative reasoning to operate autonomously in complex environments. We trace the architectural progression from statistical models to transformer-based systems, identifying capabilities that enable agentic behavior: long-range reasoning, contextual awareness, and adaptive decision-making. The chapter provides three contributions: (1) a synthesis of how LLM capabilities extend toward agency through reasoning-action-reflection loops; (2) an integrative framework describing core components perception, memory, planning, and tool execution that bridge LLMs with autonomous behavior; (3) a critical assessment of applications and persistent challenges in safety, alignment, reliability, and sustainability. Unlike existing surveys, we focus on the architectural transition from language understanding to autonomous action, emphasizing the technical gaps that must be resolved before deployment. We identify critical research priorities, including verifiable planning, scalable multi-agent coordination, persistent memory architectures, and governance frameworks. Responsible advancement requires simultaneous progress in technical robustness, interpretability, and ethical safeguards to realize potential while mitigating risks of misalignment and unintended consequences.
Why we are recommending this paper?
Due to your Interest in AI Transparency
Due to your Interest in AI Transparency
Wroclaw University of Science and Technology
AI Insights
- The Survey-Based approach demonstrated strength in predicting misclassifications for the general population, especially on consensus topics like Euthanasia and Job Guarantee. [3]
- The Hybrid model generally provided the most consistent and often highest F1 scores for predicting misclassifications across the spectrum of opinion topics. [3]
- CoDiNG: A computational model that predicts individual opinions based on their social network structure and attributes. [3]
- Survey-Based approach: Uses self-reported demographic and attribute data to predict CoDiNG model misclassifications. [3]
- Topology-Based approach: Relies exclusively on features derived from the social network structure to predict CoDiNG model misclassifications. [3]
- Hybrid model: Integrates survey-derived demographic attributes with topology-based network metrics to predict CoDiNG model misclassifications. [3]
- The study highlights the importance of considering both individual attributes and social network structure in predicting opinion model misclassifications. [3]
- The study explores how different modeling approaches (Survey-Based, Topology-Based, and Hybrid) can predict CoDiNG model misclassifications for various minority groups. [2]
- Different modeling approaches are more effective for different types of opinions (consensus, apathetic) and minority groups. [1]
Abstract
Ways in which people's opinions change are, without a doubt, subject to a rich tapestry of differing influences. Factors that affect how one arrives at an opinion reflect how they have been shaped by their environment throughout their lives, education, material status, what belief systems are they subscribed to, and what socio-economic minorities are they a part of. This already complex system is further expanded by the ever-changing nature of one's social network. It is therefore no surprise that many models have a tendency to perform best for the majority of the population and discriminating those people who are members of various marginalized groups . This bias and the study of how to counter it are subject to a rapidly developing field of Fairness in Social Network Analysis (SNA). The focus of this work is to look into how a state-of-the-art model discriminates certain minority groups and whether it is possible to reliably predict for whom it will perform worse. Moreover, is such prediction possible based solely on one's demographic or topological features? To this end, the NetSense dataset, together with a state-of-the-art CoDiNG model for opinion prediction have been employed. Our work explores how three classifier models (Demography-Based, Topology-Based, and Hybrid) perform when assessing for whom this algorithm will provide inaccurate predictions. Finally, through a comprehensive analysis of these experimental results, we identify four key patterns of algorithmic bias. Our findings suggest that no single paradigm provides the best results and that there is a real need for context-aware strategies in fairness-oriented social network analysis. We conclude that a multi-faceted approach, incorporating both individual attributes and network structures, is essential for reducing algorithmic bias and promoting inclusive decision-making.
Ways in which people's opinions change are, without a doubt, subject to a rich tapestry of differing influences. Factors that affect how one arrives at an opinion reflect how they have been shaped by their environment throughout their lives, education, material status, what belief systems are they subscribed to, and what socio-economic minorities are they a part of. This already complex system is further expanded by the ever-changing nature of one's social network. It is therefore no surprise that many models have a tendency to perform best for the majority of the population and discriminating those people who are members of various marginalized groups . This bias and the study of how to counter it are subject to a rapidly developing field of Fairness in Social Network Analysis (SNA). The focus of this work is to look into how a state-of-the-art model discriminates certain minority groups and whether it is possible to reliably predict for whom it will perform worse. Moreover, is such prediction possible based solely on one's demographic or topological features? To this end, the NetSense dataset, together with a state-of-the-art CoDiNG model for opinion prediction have been employed. Our work explores how three classifier models (Demography-Based, Topology-Based, and Hybrid) perform when assessing for whom this algorithm will provide inaccurate predictions. Finally, through a comprehensive analysis of these experimental results, we identify four key patterns of algorithmic bias. Our findings suggest that no single paradigm provides the best results and that there is a real need for context-aware strategies in fairness-oriented social network analysis. We conclude that a multi-faceted approach, incorporating both individual attributes and network structures, is essential for reducing algorithmic bias and promoting inclusive decision-making.
Why we are recommending this paper?
Due to your Interest in Data Fairness
Due to your Interest in Data Fairness
Institute for Applied Economic Research IPEA Brazil
Abstract
This paper examines the European Union's emerging regulatory landscape - focusing on the AI Act, corporate sustainability reporting and due diligence regimes (CSRD and CSDDD), and data center regulation - to assess whether it can effectively govern AI's environmental footprint. We argue that, despite incremental progress, current approaches remain ill-suited to correcting the market failures underpinning AI-related energy use, water consumption, and material demand. Key shortcomings include narrow disclosure requirements, excessive reliance on voluntary standards, weak enforcement mechanisms, and a structural disconnect between AI-specific impacts and broader sustainability laws. The analysis situates these regulatory gaps within a wider ecosystem of academic research, civil society advocacy, standard-setting, and industry initiatives, highlighting risks of regulatory capture and greenwashing. Building on this diagnosis, the paper advances strategic recommendations for the COP30 Action Agenda, calling for binding transparency obligations, harmonized international standards for lifecycle assessment, stricter governance of data center expansion, and meaningful public participation in AI infrastructure decisions.
This paper examines the European Union's emerging regulatory landscape - focusing on the AI Act, corporate sustainability reporting and due diligence regimes (CSRD and CSDDD), and data center regulation - to assess whether it can effectively govern AI's environmental footprint. We argue that, despite incremental progress, current approaches remain ill-suited to correcting the market failures underpinning AI-related energy use, water consumption, and material demand. Key shortcomings include narrow disclosure requirements, excessive reliance on voluntary standards, weak enforcement mechanisms, and a structural disconnect between AI-specific impacts and broader sustainability laws. The analysis situates these regulatory gaps within a wider ecosystem of academic research, civil society advocacy, standard-setting, and industry initiatives, highlighting risks of regulatory capture and greenwashing. Building on this diagnosis, the paper advances strategic recommendations for the COP30 Action Agenda, calling for binding transparency obligations, harmonized international standards for lifecycle assessment, stricter governance of data center expansion, and meaningful public participation in AI infrastructure decisions.
Why we are recommending this paper?
Due to your Interest in AI Ethics
Due to your Interest in AI Ethics
University of Durham
Abstract
We propose the use of a simple intuitive principle for measuring algorithmic classification bias: the significance of the differences in a classifier's error rates across the various demographics is inversely commensurate with the sample size required to statistically detect them. That is, if large sample sizes are required to statistically establish biased behavior, the algorithm is less biased, and vice versa. In a simple setting, we assume two distinct demographics, and non-parametric estimates of the error rates on them, e1 and e2, respectively. We use a well-known approximate formula for the sample size of the chi-squared test, and verify some basic desirable properties of the proposed measure. Next, we compare the proposed measure with two other commonly used statistics, the difference e2-e1 and the ratio e2/e1 of the error rates. We establish that the proposed measure is essentially different in that it can rank algorithms for bias differently, and we discuss some of its advantages over the other two measures. Finally, we briefly discuss how some of the desirable properties of the proposed measure emanate from fundamental characteristics of the method, rather than the approximate sample size formula we used, and thus, are expected to hold in more complex settings with more than two demographics.
We propose the use of a simple intuitive principle for measuring algorithmic classification bias: the significance of the differences in a classifier's error rates across the various demographics is inversely commensurate with the sample size required to statistically detect them. That is, if large sample sizes are required to statistically establish biased behavior, the algorithm is less biased, and vice versa. In a simple setting, we assume two distinct demographics, and non-parametric estimates of the error rates on them, e1 and e2, respectively. We use a well-known approximate formula for the sample size of the chi-squared test, and verify some basic desirable properties of the proposed measure. Next, we compare the proposed measure with two other commonly used statistics, the difference e2-e1 and the ratio e2/e1 of the error rates. We establish that the proposed measure is essentially different in that it can rank algorithms for bias differently, and we discuss some of its advantages over the other two measures. Finally, we briefly discuss how some of the desirable properties of the proposed measure emanate from fundamental characteristics of the method, rather than the approximate sample size formula we used, and thus, are expected to hold in more complex settings with more than two demographics.
Why we are recommending this paper?
Due to your Interest in Data Bias
Due to your Interest in Data Bias
Interests not found
We did not find any papers that match the below interests.
Try other terms also consider if the content exists in arxiv.org.
- AI Fairness
You can edit or add more interests any time.
💬 Help Shape Our Pricing
We're exploring pricing options to make this project sustainable. Take 3 minutes to share what you'd be willing to pay (if anything). Your input guides our future investment.
Help us improve your experience!
This project is on its early stages your feedback can be pivotal on the future of the project.
Let us know what you think about this week's papers and suggestions!
Give Feedback