OpenAI is internally testing a new feature called “Agora,” and it could be related to some sort of cross-platform feature that works in real time or some other new product. […] Read More
Microsoft announced on Wednesday that it disrupted RedVDS, a massive cybercrime platform linked to at least $40 million in reported losses in the United States alone since March 2025. […] Read More
Palo Alto Networks patched a high-severity vulnerability that could allow unauthenticated attackers to disable firewall protections in denial-of-service (DoS) attacks. […] Read More
Palo Alto Networks has released security updates for a high-severity security flaw impacting GlobalProtect Gateway and Portal, for which it said there exists a proof-of-concept (PoC) exploit. The vulnerability, tracked as CVE-2026-0227 (CVSS score: 7.7), has been described as a denial-of-service (DoS) condition impacting GlobalProtect PAN-OS software arising as a result of an improper check […]
Microsoft on Wednesday announced that it has taken a “coordinated legal action” in the U.S. and the U.K. to disrupt a cybercrime subscription service called RedVDS that has allegedly fueled millions in fraud losses. The effort, per the tech giant, is part of a broader law enforcement effort in collaboration with law enforcement authorities that […]
The FTC has finalized an order with General Motors, settling charges that it collected and sold the location and driving data of millions of drivers without consent. […] Read More
It’s 2026, yet many SOCs are still operating the way they did years ago, using tools and processes designed for a very different threat landscape. Given the growth in volumes and complexity of cyber threats, outdated practices no longer fully support analysts’ needs, staggering investigations and incident response. Below are four limiting habits that may […]
Experts disagree on whether the vulnerabilities in a programmable logic controller from Delta are a five-alarm fire or not much to worry over. Read More
Advancing ESG Intelligence: An Expert-level Agent and Comprehensive Benchmark for Sustainable Financecs.AI updates on arXiv.org arXiv:2601.08676v2 Announce Type: new
Abstract: Environmental, social, and governance (ESG) criteria are essential for evaluating corporate sustainability and ethical performance. However, professional ESG analysis is hindered by data fragmentation across unstructured sources, and existing large language models (LLMs) often struggle with the complex, multi-step workflows required for rigorous auditing. To address these limitations, we introduce ESGAgent, a hierarchical multi-agent system empowered by a specialized toolset, including retrieval augmentation, web search and domain-specific functions, to generate in-depth ESG analysis. Complementing this agentic system, we present a comprehensive three-level benchmark derived from 310 corporate sustainability reports, designed to evaluate capabilities ranging from atomic common-sense questions to the generation of integrated, in-depth analysis. Empirical evaluations demonstrate that ESGAgent outperforms state-of-the-art closed-source LLMs with an average accuracy of 84.15% on atomic question-answering tasks, and excels in professional report generation by integrating rich charts and verifiable references. These findings confirm the diagnostic value of our benchmark, establishing it as a vital testbed for assessing general and advanced agentic capabilities in high-stakes vertical domains.
arXiv:2601.08676v2 Announce Type: new
Abstract: Environmental, social, and governance (ESG) criteria are essential for evaluating corporate sustainability and ethical performance. However, professional ESG analysis is hindered by data fragmentation across unstructured sources, and existing large language models (LLMs) often struggle with the complex, multi-step workflows required for rigorous auditing. To address these limitations, we introduce ESGAgent, a hierarchical multi-agent system empowered by a specialized toolset, including retrieval augmentation, web search and domain-specific functions, to generate in-depth ESG analysis. Complementing this agentic system, we present a comprehensive three-level benchmark derived from 310 corporate sustainability reports, designed to evaluate capabilities ranging from atomic common-sense questions to the generation of integrated, in-depth analysis. Empirical evaluations demonstrate that ESGAgent outperforms state-of-the-art closed-source LLMs with an average accuracy of 84.15% on atomic question-answering tasks, and excels in professional report generation by integrating rich charts and verifiable references. These findings confirm the diagnostic value of our benchmark, establishing it as a vital testbed for assessing general and advanced agentic capabilities in high-stakes vertical domains. Read More
Evaluating the Ability of Explanations to Disambiguate Models in a Rashomon Setcs.AI updates on arXiv.org arXiv:2601.08703v1 Announce Type: new
Abstract: Explainable artificial intelligence (XAI) is concerned with producing explanations indicating the inner workings of models. For a Rashomon set of similarly performing models, explanations provide a way of disambiguating the behavior of individual models, helping select models for deployment. However explanations themselves can vary depending on the explainer used, and need to be evaluated. In the paper “Evaluating Model Explanations without Ground Truth”, we proposed three principles of explanation evaluation and a new method “AXE” to evaluate the quality of feature-importance explanations. We go on to illustrate how evaluation metrics that rely on comparing model explanations against ideal ground truth explanations obscure behavioral differences within a Rashomon set. Explanation evaluation aligned with our proposed principles would highlight these differences instead, helping select models from the Rashomon set. The selection of alternate models from the Rashomon set can maintain identical predictions but mislead explainers into generating false explanations, and mislead evaluation methods into considering the false explanations to be of high quality. AXE, our proposed explanation evaluation method, can detect this adversarial fairwashing of explanations with a 100% success rate. Unlike prior explanation evaluation strategies such as those based on model sensitivity or ground truth comparison, AXE can determine when protected attributes are used to make predictions.
arXiv:2601.08703v1 Announce Type: new
Abstract: Explainable artificial intelligence (XAI) is concerned with producing explanations indicating the inner workings of models. For a Rashomon set of similarly performing models, explanations provide a way of disambiguating the behavior of individual models, helping select models for deployment. However explanations themselves can vary depending on the explainer used, and need to be evaluated. In the paper “Evaluating Model Explanations without Ground Truth”, we proposed three principles of explanation evaluation and a new method “AXE” to evaluate the quality of feature-importance explanations. We go on to illustrate how evaluation metrics that rely on comparing model explanations against ideal ground truth explanations obscure behavioral differences within a Rashomon set. Explanation evaluation aligned with our proposed principles would highlight these differences instead, helping select models from the Rashomon set. The selection of alternate models from the Rashomon set can maintain identical predictions but mislead explainers into generating false explanations, and mislead evaluation methods into considering the false explanations to be of high quality. AXE, our proposed explanation evaluation method, can detect this adversarial fairwashing of explanations with a 100% success rate. Unlike prior explanation evaluation strategies such as those based on model sensitivity or ground truth comparison, AXE can determine when protected attributes are used to make predictions. Read More
