Artificial intelligence has been a tricky subject in Washington. Most lawmakers agree that it poses significant dangers if left unregulated, yet there remains a lack of consensus on how to tackle these concerns. But speaking at a TIME100 Talks conversation on Friday ahead of the White House Correspondents Dinner, a panel of experts with backgrounds in government, national security, and social justice expressed optimism that the U.S. government will finally "get it right" so that society can reap the benefits of AI while safeguarding against potential dangers. "We can't afford to get this wrong--again," Shalanda Young, the director of the Office of Management and Budget in the Biden Administration, told TIME Senior White House Correspondent Brian Bennett. "The government was already behind the tech boom. Can you imagine if the government is a user of AI and we get that wrong?"

Relation extraction (RE) aims to identify relations between entities mentioned in texts. Although large language models (LLMs) have demonstrated impressive in-context learning (ICL) abilities in various tasks, they still suffer from poor performances compared to most supervised fine-tuned RE methods. Utilizing ICL for RE with LLMs encounters two challenges: (1) retrieving good demonstrations from training examples, and (2) enabling LLMs exhibit strong ICL abilities in RE. On the one hand, retrieving good demonstrations is a non-trivial process in RE, which easily results in low relevance regarding entities and relations. On the other hand, ICL with an LLM achieves poor performance in RE while RE is different from language modeling in nature or the LLM is not large enough. In this work, we propose a novel recall-retrieve-reason RE framework that synergizes LLMs with retrieval corpora (training examples) to enable relevant retrieving and reliable in-context reasoning. Specifically, we distill the consistently ontological knowledge from training datasets to let LLMs generate relevant entity pairs grounded by retrieval corpora as valid queries. These entity pairs are then used to retrieve relevant training examples from the retrieval corpora as demonstrations for LLMs to conduct better ICL via instruction tuning. Extensive experiments on different LLMs and RE datasets demonstrate that our method generates relevant and valid entity pairs and boosts ICL abilities of LLMs, achieving competitive or new state-of-the-art performance on sentence-level RE compared to previous supervised fine-tuning methods and ICL-based methods.

Cybersecurity is one of the global issues because of the extensive dependence on cyber systems of individuals, industries, and organizations. Among the cyber attacks, phishing is increasing tremendously and affecting the global economy. Therefore, this phenomenon highlights the vital need for enhancing user awareness and robust support at both individual and organizational levels. Phishing URL identification is the best way to address the problem. Various machine learning and deep learning methods have been proposed to automate the detection of phishing URLs. However, these approaches often need more convincing accuracy and rely on datasets consisting of limited samples. Furthermore, these black box intelligent models decision to detect suspicious URLs needs proper explanation to understand the features affecting the output. To address the issues, we propose a 1D Convolutional Neural Network (CNN) and trained the model with extensive features and a substantial amount of data. The proposed model outperforms existing works by attaining an accuracy of 99.85%. Additionally, our explainability analysis highlights certain features that significantly contribute to identifying the phishing URL.

In recent years, software systems powered by deep learning (DL) techniques have significantly facilitated people's lives in many aspects. As the backbone of these DL systems, various DL libraries undertake the underlying optimization and computation. However, like traditional software, DL libraries are not immune to bugs, which can pose serious threats to users' personal property and safety. Studying the characteristics of DL libraries, their associated bugs, and the corresponding testing methods is crucial for enhancing the security of DL systems and advancing the widespread application of DL technology. This paper provides an overview of the testing research related to various DL libraries, discusses the strengths and weaknesses of existing methods, and provides guidance and reference for the application of the DL library. This paper first introduces the workflow of DL underlying libraries and the characteristics of three kinds of DL libraries involved, namely DL framework, DL compiler, and DL hardware library. It then provides definitions for DL underlying library bugs and testing. Additionally, this paper summarizes the existing testing methods and tools tailored to these DL libraries separately and analyzes their effectiveness and limitations. It also discusses the existing challenges of DL library testing and outlines potential directions for future research.

Graph Neural Networks (GNNs) have demonstrated state-of-the-art performance in various graph representation learning tasks. Recently, studies revealed their vulnerability to adversarial attacks. In this work, we theoretically define the concept of expected robustness in the context of attributed graphs and relate it to the classical definition of adversarial robustness in the graph representation learning literature. Our definition allows us to derive an upper bound of the expected robustness of Graph Convolutional Networks (GCNs) and Graph Isomorphism Networks subject to node feature attacks. Building on these findings, we connect the expected robustness of GNNs to the orthonormality of their weight matrices and consequently propose an attack-independent, more robust variant of the GCN, called the Graph Convolutional Orthonormal Robust Networks (GCORNs). We further introduce a probabilistic method to estimate the expected robustness, which allows us to evaluate the effectiveness of GCORN on several real-world datasets. Experimental experiments showed that GCORN outperforms available defense methods. Our code is publicly available at: https://github.com/Sennadir/GCORN. Graph-structured data is prevalent in a wide range of domains, motivating therefore the development of neural network models that can operate on graphs, known as Graph Neural Networks (GNNs). GNNs have emerged as a powerful tool for learning node and graph representations. Many GNNs are instances of Message Passing Neural Networks (MPNNs) (Gilmer et al., 2017) such as Graph Isomorphism Networks (GIN)(Xu et al., 2019b) and Graph Convolutional Networks (GCN)(Kipf & Welling, 2017). These models have been successfully applied in real-world applications such as molecular design (Kearnes et al., 2016). In parallel to their success, it has been shown, particularly in the field of computer vision, that deep learning architectures can be susceptible to adversarial attacks (Goodfellow et al., 2015).

Perceptions of hate can vary greatly across cultural contexts. Hate speech (HS) datasets, however, have traditionally been developed by language. This hides potential cultural biases, as one language may be spoken in different countries home to different cultures. In this work, we evaluate cultural bias in HS datasets by leveraging two interrelated cultural proxies: language and geography. We conduct a systematic survey of HS datasets in eight languages and confirm past findings on their English-language bias, but also show that this bias has been steadily decreasing in the past few years. For three geographically-widespread languages -- English, Arabic and Spanish -- we then leverage geographical metadata from tweets to approximate geo-cultural contexts by pairing language and country information. We find that HS datasets for these languages exhibit a strong geo-cultural bias, largely overrepresenting a handful of countries (e.g., US and UK for English) relative to their prominence in both the broader social media population and the general population speaking these languages. Based on these findings, we formulate recommendations for the creation of future HS datasets.

Iranian-backed Houthis rebels have claimed responsibility for a U.S. MQ-9 Reaper drone crash off the coast of Yemen on Thursday, Fox News confirmed on Friday. Thursday's crash is the fourth remotely piloted drone brought down by Iranian-proxy groups since November, costing the U.S. government upwards of 120 million. It is also the third time Houthi rebels have brought down a U.S. MQ-9 drone. Remotely piloted MQ-9 Reaper drones cost around 30 million each. Last fall, the Houthis released video of a reaper drone the rebels shot down on Nov. 8, one day after Hamas' unprovoked attack on Israel.

Fox News Flash top headlines are here. Check out what's clicking on Foxnews.com. Pope Francis is taking his call for artificial intelligence to be developed and used according to ethical lines to the Group of 7 industrialized nations. Italian Prime Minister Giorgia Meloni announced Friday that Francis had accepted her invitation to attend the G7 Summit in Puglia in June. The Vatican confirmed the news.

A federal report published today found that Tesla's Autopilot system was involved in at least 13 fatal crashes in which drivers misused the system in ways the automaker should have foreseen--and done more to prevent. Not only that, but the report called out Tesla as an "industry outlier" because its driver assistance features lacked some of the basic precautions taken by its competitors. Now regulators are questioning whether a Tesla Autopilot update designed to fix these basic design issues and prevent fatal incidents has gone far enough. These fatal crashes killed 14 people and injured 49, according to data collected and published by the National Highway Traffic Safety Administration, the federal road-safety regulator in the US. At least half of the 109 "frontal plane" crashes closely examined by government engineers--those in which a Tesla crashed into a vehicle or obstacle directly in its path--involved hazards visible five seconds or more before impact.

US auto-safety regulators said on Friday that their investigation into Tesla's Autopilot had identified at least 13 fatal crashes in which the feature had been involved. The investigation also found the electric carmaker's claims did not match up with reality. The National Highway Traffic Safety Administration (NHTSA) disclosed on Friday that during its three-year Autopilot safety investigation, which it launched in August 2021, it identified at least 13 Tesla crashes involving one or more death, and many more involving serious injuries, in which "foreseeable driver misuse of the system played an apparent role". It also found evidence that "Tesla's weak driver engagement system was not appropriate for Autopilot's permissive operating capabilities", which resulted in a "critical safety gap". The NHTSA also raised concerns that Tesla's Autopilot name "may lead drivers to believe that the automation has greater capabilities than it does and invite drivers to overly trust the automation".