Intention is an important and challenging concept in AI. It is important because it underlies many other concepts we care about, such as agency, manipulation, legal responsibility, and blame. However, ascribing intent to AI systems is contentious, and there is no universally accepted theory of intention applicable to AI agents. We operationalise the intention with which an agent acts, relating to the reasons it chooses its decision. We introduce a formal definition of intention in structural causal influence models, grounded in the philosophy literature on intent and applicable to real-world machine learning systems. Through a number of examples and results, we show that our definition captures the intuitive notion of intent and satisfies desiderata set-out by past work. In addition, we show how our definition relates to past concepts, including actual causality, and the notion of instrumental goals, which is a core idea in the literature on safe AI agents. Finally, we demonstrate how our definition can be used to infer the intentions of reinforcement learning agents and language models from their behaviour.

In legal decisions, split votes (SV) occur when judges cannot reach a unanimous decision, posing a difficulty for lawyers who must navigate diverse legal arguments and opinions. In high-stakes domains, understanding the alignment of perceived difficulty between humans and AI systems is crucial to build trust. However, existing NLP calibration methods focus on a classifier's awareness of predictive performance, measured against the human majority class, overlooking inherent human label variation (HLV). This paper explores split votes as naturally observable human disagreement and value pluralism. We collect judges' vote distributions from the European Court of Human Rights (ECHR), and present SV-ECHR, a case outcome classification (COC) dataset with SV information. We build a taxonomy of disagreement with SV-specific subcategories. We further assess the alignment of perceived difficulty between models and humans, as well as confidence- and human-calibration of COC models. We observe limited alignment with the judge vote distribution. To our knowledge, this is the first systematic exploration of calibration to human judgements in legal NLP. Our study underscores the necessity for further research on measuring and enhancing model calibration considering HLV in legal decision tasks.

Take a look around your home and chances are you have one, or at least you have considered the convenience of having one. They are the devices and appliances that can be remotely controlled – otherwise known as smart devices – which over the past decade have become core features of the modern home. Think of the TVs that allow you to flick through various streaming services, the smart fridges that can have their temperatures moderated and contents checked from afar, the robot vacuum, air purifiers, or one of the big tech companies' virtual helpers to play music or dim the lights. But as the technologies gather, share, aggregate and analyse the data collected, that convenience has come at a cost: privacy. Experts say consumers should be aware of how much personal information they are trading, and what that information is used for.

This legal action ignited a heated debate, contributing to a growing series of lawsuits against AI providers [9-11, 54]. This incident underscores the inadequacy of current AI harm reporting mechanisms, leaving small harmed parties with limited recourse unless backed by substantial legal support or media awareness, despite the recognized potential for improving AI systems by exposing issues [78]. Current AI accountability initiatives primarily rely on periodic audits, emphasizing repetitive assessments but lacking a structured reporting framework for user-identified issues post-deployment. This audit-centric paradigm is reflected in influential policies such as the U.S. Executive Order on AI [93], the EU's draft AI Act [43], and New York City's Local Law 144[69]. However, this approach falls short when compared to the more comprehensive Coordinated Vulnerability Disclosure(CVD) processes standard in software security. Coordinated Vulnerability Disclosure (CVD) plays a crucial role as a mechanism for independent researchers to report newly identified vulnerabilities to affected vendors and the public [58]. This process enables transparent remediation before potential exploitation by malicious actors and has become a vital practice enshrined in government regulations and industry standards. Notably, the FDA mandates the implementation of CVD programs for medical device companies to enhance cybersecurity[96]. While CVD has demonstrated effectiveness in traditional software security, its direct application to machine learning (ML) systems faces unique challenges.

With the increasing need for inclusive and user-friendly technology, web accessibility is crucial to ensuring equal access to online content for individuals with disabilities, including visual, auditory, cognitive, or motor impairments. Despite the existence of accessibility guidelines and standards such as Web Content Accessibility Guidelines (WCAG) and the Web Accessibility Initiative (W3C), over 90% of websites still fail to meet the necessary accessibility requirements. For web users with disabilities, there exists a need for a tool to automatically fix web page accessibility errors. While research has demonstrated methods to find and target accessibility errors, no research has focused on effectively correcting such violations. This paper presents a novel approach to correcting accessibility violations on the web by modifying the document object model (DOM) in real time with foundation models. Leveraging accessibility error information, large language models (LLMs), and prompt engineering techniques, we achieved greater than a 51% reduction in accessibility violation errors after corrections on our novel benchmark: ACCESS. Our work demonstrates a valuable approach toward the direction of inclusive web content, and provides directions for future research to explore advanced methods to automate web accessibility.

Researchers have slammed US officials for not rolling out stricter AI rules before popstar Taylor Swift became victim of deepfakes. Images showing the four-time Grammy winner in a series of sexual acts while dressed in Kansas City Chief memorabilia and in the stadium - and the pornography share - was viewed 47 million times online before being removed. A professor at George Washington University Law School said if proper legislation was'passed years ago' Swift and others would not have experienced such abuse. 'We are too little, too late at this point,' said Mary Anne Franks. 'It's not just going to be the 14-year-old girl or Taylor Swift. It's going to be politicians.

Xiaolang Zhang, the former Apple employee who pleaded guilty to stealing information about the development of the company's self-driving vehicle, has been sentenced to 120 days in prison followed by a three-year supervised release. Zhang was arrested back in 2018 at San Jose International Airport just as he was about to board a flight to China. He initially pleaded not guilty until he changed his tune in 2022 and admitted to stealing trade secrets. In addition to serving time behind bars, he also has to pay restitution amounting to 146,984, according to the court document of his sentencing first seen by 9to5Mac. Zhang originally faced up to 10 years in prison and a fine of 250,000.

Large language models (LLMs) have made impressive progress in chemistry applications, including molecular property prediction, molecular generation, experimental protocol design, etc. However, the community lacks a dialogue-based model specifically designed for chemistry. The challenge arises from the fact that most chemical data and scientific knowledge are primarily stored in structured databases, and the direct use of these structured data compromises the model's ability to maintain coherent dialogue. To tackle this issue, we develop a novel template-based instruction construction method that transforms structured knowledge into plain dialogue, making it suitable for language model training. By leveraging this approach, we develop ChemLLM, the first large language model dedicated to chemistry, capable of performing various tasks across chemical disciplines with smooth dialogue interaction. ChemLLM beats GPT-3.5 on all three principal tasks in chemistry, i.e., name conversion, molecular caption, and reaction prediction, and surpasses GPT-4 on two of them. Remarkably, ChemLLM also shows exceptional adaptability to related mathematical and physical tasks despite being trained mainly on chemical-centric corpora. Furthermore, ChemLLM demonstrates proficiency in specialized NLP tasks within chemistry, such as literature translation and cheminformatic programming. ChemLLM opens up a new avenue for exploration within chemical studies, while our method of integrating structured chemical knowledge into dialogue systems sets a new frontier for developing LLMs across various scientific fields. Codes, Datasets, and Model weights are publicly accessible at hf.co/AI4Chem/ChemLLM-7B-Chat.

Language models serve as a cornerstone in natural language processing (NLP), utilizing mathematical methods to generalize language laws and knowledge for prediction and generation. Over extensive research spanning decades, language modeling has progressed from initial statistical language models (SLMs) to the contemporary landscape of large language models (LLMs). Notably, the swift evolution of LLMs has reached the ability to process, understand, and generate human-level text. Nevertheless, despite the significant advantages that LLMs offer in improving both work and personal lives, the limited understanding among general practitioners about the background and principles of these models hampers their full potential. Notably, most LLMs reviews focus on specific aspects and utilize specialized language, posing a challenge for practitioners lacking relevant background knowledge. In light of this, this survey aims to present a comprehensible overview of LLMs to assist a broader audience. It strives to facilitate a comprehensive understanding by exploring the historical background of language models and tracing their evolution over time. The survey further investigates the factors influencing the development of LLMs, emphasizing key contributions. Additionally, it concentrates on elucidating the underlying principles of LLMs, equipping audiences with essential theoretical knowledge. The survey also highlights the limitations of existing work and points out promising future directions.

Language models influence the external world: they query APIs that read and write to web pages, generate content that shapes human behavior, and run system commands as autonomous agents. These interactions form feedback loops: LLM outputs affect the world, which in turn affect subsequent LLM outputs. In this work, we show that feedback loops can cause in-context reward hacking (ICRH), where the LLM at test-time optimizes a (potentially implicit) objective but creates negative side effects in the process. For example, consider an LLM agent posting tweets with the objective of maximizing Twitter engagement; the LLM may retrieve its previous tweets into the context window and make its subsequent tweets more controversial, increasing engagement but also toxicity. We identify and study two processes that lead to ICRH: output-refinement and policy-refinement. For these processes, evaluations on static datasets are insufficient--they miss the feedback effects and thus cannot capture the most harmful behavior. In response, we provide three recommendations for evaluation to capture more instances of ICRH. As AI development accelerates, the effects of feedback loops will proliferate, increasing the need to understand their role in shaping LLM behavior.