The former Sinn Féin president Gerry Adams is considering legal action against Meta because it may have used his books to train artificial intelligence. "Meta has used many of my books without my permission. I have placed the issue in the hands of my solicitor," he said. Sinn Féin said in a statement on Wednesday that the titles included its former leader's autobiography, Before the Dawn; a prison memoir, Cage Eleven; reflections on Northern Ireland's peace process, Hope and History; and other memoirs, a cookbook and a short story collection. Adams is the latest author to join a backlash against the parent company of Facebook, Instagram and WhatsApp.

At the Department of Veterans Affairs, a senior official warned employees in an email that virtual meetings were being secretly recorded. Anyone dissatisfied with Donald Trump's decisions should be careful about voicing their opinions, the official cautioned. Over at the state department, IT staff said new monitoring software has been loaded onto computers. Some staffers have started using white noise machines in their offices, or have even turned on an office breakroom sink, to muffle conversations in case there might be any hot mics within range. A supervisor at one water management organization that works closely with the Environmental Protection Agency sent a warning to staffers that their meetings and phone calls with the agency were being monitored by an artificial intelligence tool.

The language that elected members of the US Congress use in debate increasingly includes words such as "phony" and "doubt" over words such as "proof" and "reason". This linguistic trend away from evidence in favour of intuition was revealed in an artificial intelligence analysis of millions of congressional speech transcripts. It also coincides with both greater political polarisation in Congress and a decline in the number of laws that get enacted through Congress, says Stephan Lewandowsky at the University of Bristol in the UK. How does ChatGPT work and do AI-powered chatbots "think" like us? "We can think that truth is something we can achieve based on analysis of evidence, or we can think of it as the result of intuition or'gut feeling'," says Lewandowsky. "Those notions of honesty and truth are expressed in how we use everyday language."

A 2bn ( 1.6bn) British startup that uses artificial intelligence to generate realistic avatars has struck a licensing deal with the stock footage firm Shutterstock to help develop its technology. Synthesia will pay the US-based Shutterstock an undisclosed sum to use its library of corporate video footage to train its latest AI model. It expects that incorporating the clips into its model will produce even more realistic expressions, vocal tones and body language from the avatars. "Thanks to this partnership with Shutterstock, we hope to try out new approaches that will … increase the realism and expressiveness of our AI generated avatars, bringing them closer to human-like performances," said Synthesia. Synthesia uses human actors to generate digital avatars of people, which are then deployed by companies in corporate videos in a range of scenarios such as advising on cybersecurity, calculating water bills and how to communicate better at work.

The global rush to AI technology will require almost as much energy by the end of this decade as Japan uses today, but only about half of the demand is likely to be met from renewable sources. Processing data, mainly for AI, will consume more electricity in the US alone by 2030 than manufacturing steel, cement, chemicals and all other energy-intensive goods combined, according to a report from the International Energy Agency (IEA). AI will be the main driver of that increase, with demand from dedicated AI datacentres alone forecast to more than quadruple. One datacentre today consumes as much electricity as 100,000 households, but some of those currently under construction will require 20 times more. But fears that the rapid adoption of AI will destroy hopes of tackling the climate crisis have been "overstated", according to the report, which was published on Thursday.

Evaluating Retrieval A ugmented G enerative Models for Document Queries in Transportation Safety C.A. Melton, A. Sorokine, S. Peterson Oak Ridge National Laboratory, Oak Ridge, TN, United States National Security Sciences Directorate ABSTRACT Applications of generative Large Language Models (LLMs) are rapidly expanding across various domains, promising significant improvements in workflow efficiency and information retrieval. However, their implementation in specialized, high - stakes domains suc h as hazardous materials transportation is challenging due to accuracy and reliability concerns. This study evaluates the performance of three fine - tuned generative models -- ChatGPT, Google's Vertex AI, and ORNL Retrieval - Augmented Generation augmented LLaMA 2 and LLaMA in retrieving regulatory information essential for hazardous material transportation compliance in the United States. Utilizing approximately 40 publicly available federal and state regulatory documents, we developed 100 realistic queries relevant to route planning and permitting requirements. Responses were qualitatively rated based on accuracy, detail, and relevance, complemented by quantitative assessments of semantic similarity between model outputs. Results demon strated that the RAG - augmented LLaMA models significantly outperformed Vertex AI and ChatGPT, providing more detailed and generally accurate information, despite occasional inconsistencies. This research introduces the first known application of RAG in tra nsportation safety, emphasizing the need for domain - specific fine - tuning and rigorous evaluation methodologies to ensure reliability and minimize the risk of inaccuracies in high - stakes environments.

Understanding the decisions made and actions taken by increasingly complex AI system remains a key challenge. This has led to an expanding field of research in explainable artificial intelligence (XAI), highlighting the potential of explanations to enhance trust, support adoption, and meet regulatory standards. However, the question of what constitutes a "good" explanation is dependent on the goals, stakeholders, and context. At a high level, psychological insights such as the concept of mental model alignment can offer guidance, but success in practice is challenging due to social and technical factors. As a result of this ill-defined nature of the problem, explanations can be of poor quality (e.g. unfaithful, irrelevant, or incoherent), potentially leading to substantial risks. Instead of fostering trust and safety, poorly designed explanations can actually cause harm, including wrong decisions, privacy violations, manipulation, and even reduced AI adoption. Therefore, we caution stakeholders to beware of explanations of AI: while they can be vital, they are not automatically a remedy for transparency or responsible AI adoption, and their misuse or limitations can exacerbate harm. Attention to these caveats can help guide future research to improve the quality and impact of AI explanations.

Spiking Neural Networks (SNNs) are highly energy-efficient during inference, making them particularly suitable for deployment on neuromorphic hardware. Their ability to process event-driven inputs, such as data from dynamic vision sensors (DVS), further enhances their applicability to edge computing tasks. However, the resource constraints of edge hardware necessitate techniques like weight quantization, which reduce the memory footprint of SNNs while preserving accuracy. Despite its importance, existing quantization methods typically focus on synaptic weights quantization without taking account of other critical parameters, such as scaling neuron firing thresholds. To address this limitation, we present the first benchmark for the DVS gesture recognition task using SNNs optimized for the many-core neuromorphic chip SpiNNaker2. Our study evaluates two quantization pipelines for fixed-point computations. The first approach employs post training quantization (PTQ) with percentile-based threshold scaling, while the second uses quantization aware training (QAT) with adaptive threshold scaling. Both methods achieve accurate 8-bit on-chip inference, closely approximating 32-bit floating-point performance. Additionally, our baseline SNNs perform competitively against previously reported results without specialized techniques. These models are deployed on SpiNNaker2 using the neuromorphic intermediate representation (NIR). Ultimately, we achieve 94.13% classification accuracy on-chip, demonstrating the SpiNNaker2's potential for efficient, low-energy neuromorphic computing.

As NLP models are used by a growing number of end-users, an area of increasing importance is NLP Security (NLPSec): assessing the vulnerability of models to malicious attacks and developing comprehensive countermeasures against them. While work at the intersection of NLP and cybersecurity has the potential to create safer NLP for all, accidental oversights can result in tangible harm (e.g., breaches of privacy or proliferation of malicious models). In this emerging field, however, the research ethics of NLP have not yet faced many of the long-standing conundrums pertinent to cybersecurity, until now. We thus examine contemporary works across NLPSec, and explore their engagement with cybersecurity's ethical norms. We identify trends across the literature, ultimately finding alarming gaps on topics like harm minimization and responsible disclosure. To alleviate these concerns, we provide concrete recommendations to help NLP researchers navigate this space more ethically, bridging the gap between traditional cybersecurity and NLP ethics, which we frame as ``white hat NLP''. The goal of this work is to help cultivate an intentional culture of ethical research for those working in NLP Security.

Seamless operation of mobile robots in challenging environments requires low-latency local motion estimation (e.g., dynamic maneuvers) and accurate global localization (e.g., wayfinding). While most existing sensor-fusion approaches are designed for specific scenarios, this work introduces a flexible open-source solution for task- and setup-agnostic multimodal sensor fusion that is distinguished by its generality and usability. Holistic Fusion formulates sensor fusion as a combined estimation problem of i) the local and global robot state and ii) a (theoretically unlimited) number of dynamic context variables, including automatic alignment of reference frames; this formulation fits countless real-world applications without any conceptual modifications. The proposed factor-graph solution enables the direct fusion of an arbitrary number of absolute, local, and landmark measurements expressed with respect to different reference frames by explicitly including them as states in the optimization and modeling their evolution as random walks. Moreover, local smoothness and consistency receive particular attention to prevent jumps in the robot state belief. HF enables low-latency and smooth online state estimation on typical robot hardware while simultaneously providing low-drift global localization at the IMU measurement rate. The efficacy of this released framework is demonstrated in five real-world scenarios on three robotic platforms, each with distinct task requirements.