Large language models (LLMs) have recently revolutionized language processing tasks but have also brought ethical and legal issues. LLMs have a tendency to memorize potentially private or copyrighted information present in the training data, which might then be delivered to end users at inference time. When this happens, a naive solution is to retrain the model from scratch after excluding the undesired data. Although this guarantees that the target data have been forgotten, it is also prohibitively expensive for LLMs. Approximate unlearning offers a more efficient alternative, as it consists of ex post modifications of the trained model itself to prevent undesirable results, but it lacks forgetting guarantees because it relies solely on empirical evidence. In this work, we present DP2Unlearning, a novel LLM unlearning framework that offers formal forgetting guarantees at a significantly lower cost than retraining from scratch on the data to be retained. DP2Unlearning involves training LLMs on textual data protected using ε-differential privacy (DP), which later enables efficient unlearning with the guarantees against disclosure associated with the chosen ε. Our experiments demonstrate that DP2Unlearning achieves similar model performance post-unlearning, compared to an LLM retraining from scratch on retained data -- the gold standard exact unlearning -- but at approximately half the unlearning cost. In addition, with a reasonable computational cost, it outperforms approximate unlearning methods at both preserving the utility of the model post-unlearning and effectively forgetting the targeted information.

As the old adage goes, "On the internet, nobody knows you're a dog". But in Australia it might soon be the case that everything from search engines and social media sites, to app stores and AI chatbots will have to know your age. The Albanese government trumpeted the passage of its legislation banning under 16s from social media – which will come into effect in December – but new industry codes developed by the tech sector and eSafety commissioner Julie Inman Grant under the Online Safety Act will probably have much larger ramifications for how Australians access the internet. Measures to be deployed by online services could include looking at your account history, or using facial age assurance and bank card checks. Identity checks using IDs such as drivers licences to keep children under 16 off social media will also apply to logged-in accounts for search engines from December, under an industry code that came into force at the end of June.

Developing precise and computationally efficient traffic accident anticipation system is crucial for contemporary autonomous driving technologies, enabling timely intervention and loss prevention. In this paper, we propose an accident anticipation framework employing a dual-branch architecture that effectively integrates visual information from dashcam videos with structured textual data derived from accident reports. Furthermore, we introduce a feature aggregation method that facilitates seamless integration of multimodal inputs through large models (GPT-4o, Long-CLIP), complemented by targeted prompt engineering strategies to produce actionable feedback and standardized accident archives. Comprehensive evaluations conducted on benchmark datasets (DAD, CCD, and A3D) validate the superior predictive accuracy, enhanced responsiveness, reduced computational overhead, and improved interpretability of our approach, thus establishing a new benchmark for state-of-the-art performance in traffic accident anticipation.

We study the problem of clustering ranking vectors, where each vector represents preferences as an ordered list of distinct integers. Specifically, we focus on the k-centroids ranking vectors clustering problem (KRC), which aims to partition a set of ranking vectors into k clusters and identify the centroid of each cluster. Unlike classical k-means clustering (KMC), KRC constrains both the observations and centroids to be ranking vectors. We establish the NP-hardness of KRC and characterize its feasible set. For the single-cluster case, we derive a closed-form analytical solution for the optimal centroid, which can be computed in linear time. To address the computational challenges of KRC, we develop an efficient approximation algorithm, KRCA, which iteratively refines initial solutions from KMC, referred to as the baseline solution. Additionally, we introduce a branch-and-bound (BnB) algorithm for efficient cluster reconstruction within KRCA, leveraging a decision tree framework to reduce computational time while incorporating a controlling parameter to balance solution quality and efficiency. We establish theoretical error bounds for KRCA and BnB. Through extensive numerical experiments on synthetic and real-world datasets, we demonstrate that KRCA consistently outperforms baseline solutions, delivering significant improvements in solution quality with fast computational times. This work highlights the practical significance of KRC for personalization and large-scale decision making, offering methodological advancements and insights that can be built upon in future studies.

The increasing use of generative AI for resume screening is predicated on the assumption that it offers an unbiased alternative to biased human decision-making. However, this belief fails to address a critical question: are these AI systems fundamentally competent at the evaluative tasks they are meant to perform? This study investigates the question of competence through a two-part audit of eight major AI platforms. Experiment 1 confirmed complex, contextual racial and gender biases, with some models penalizing candidates merely for the presence of demographic signals. Experiment 2, which evaluated core competence, provided a critical insight: some models that appeared unbiased were, in fact, incapable of performing a substantive evaluation, relying instead on superficial keyword matching. This paper introduces the "Illusion of Neutrality" to describe this phenomenon, where an apparent lack of bias is merely a symptom of a model's inability to make meaningful judgments. This study recommends that organizations and regulators adopt a dual-validation framework, auditing AI hiring tools for both demographic bias and demonstrable competence to ensure they are both equitable and effective.

Tabular data is one of the most prevalent and important data formats in real-world applications such as healthcare, finance, and education. However, its effective use in machine learning is often constrained by data scarcity, privacy concerns, and class imbalance. Synthetic tabular data generation has emerged as a powerful solution, leveraging generative models to learn underlying data distributions and produce realistic, privacy-preserving samples. Although this area has seen growing attention, most existing surveys focus narrowly on specific methods (e.g., GANs or privacy-enhancing techniques), lacking a unified and comprehensive view that integrates recent advances such as diffusion models and large language models (LLMs). In this survey, we present a structured and in-depth review of synthetic tabular data generation methods. Specifically, the survey is organized into three core components: (1) Background, which covers the overall generation pipeline, including problem definitions, synthetic tabular data generation methods, post processing, and evaluation; (2) Generation Methods, where we categorize existing approaches into traditional generation methods, diffusion model methods, and LLM-based methods, and compare them in terms of architecture, generation quality, and applicability; and (3) Applications and Challenges, which summarizes practical use cases, highlights common datasets, and discusses open challenges such as heterogeneity, data fidelity, and privacy protection. This survey aims to provide researchers and practitioners with a holistic understanding of the field and to highlight key directions for future work in synthetic tabular data generation.

Multimodal large language models (MLLMs) excel in vision-language tasks but also pose significant risks of generating harmful content, particularly through jailbreak attacks. Jailbreak attacks refer to intentional manipulations that bypass safety mechanisms in models, leading to the generation of inappropriate or unsafe content. Detecting such attacks is critical to ensuring the responsible deployment of MLLMs. Existing jailbreak detection methods face three primary challenges: (1) Many rely on model hidden states or gradients, limiting their applicability to white-box models, where the internal workings of the model are accessible; (2) They involve high computational overhead from uncertainty-based analysis, which limits real-time detection, and (3) They require fully labeled harmful datasets, which are often scarce in real-world settings. To address these issues, we introduce a test-time adaptive framework called JAILDAM. Our method leverages a memory-based approach guided by policy-driven unsafe knowledge representations, eliminating the need for explicit exposure to harmful data. By dynamically updating unsafe knowledge during test-time, our framework improves generalization to unseen jailbreak strategies while maintaining efficiency. Experiments on multiple VLM jailbreak benchmarks demonstrate that JAILDAM delivers state-of-the-art performance in harmful content detection, improving both accuracy and speed.

Roblox is rolling out new features aimed at making the platform safer for minors, including a revamped friend system, privacy tools, and age verification services users submit by recording a video selfie. In Roblox's old friend system, players have no distinction between people they know casually or online versus someone they consider a close friend. The platform's new tiered system introduces Connections and Trusted Connections specifically for people that players know and trust. To access Trusted Connections and its benefits, users first need to complete an age verification, which requires them to submit a video selfie. Once they've submitted their video, the company says it's run against an AI-driven "diverse dataset" to get an age estimation.

Getting into local models takes a bit more effort than, say, navigating to ChatGPT's online interface. But the very accessibility of a tool like ChatGPT comes with a cost. "It's the classic adage: If something's free, you're the product," says Elizabeth Seger, the director of digital policy at Demos, a London-based think tank. OpenAI, which offers both paid and free tiers, trains its models on users' chats by default. It's not too difficult to opt out of this training, and it also used to be possible to remove your chat data from OpenAI's systems entirely, until a recent legal decision in the New York Times' ongoing lawsuit against OpenAI required the company to maintain all user conversations with ChatGPT.

Fox News anchor Bret Baier examines the U.S. power supply on'Special Report.' The global race to harness the power of artificial intelligence (AI) has begun. President Donald Trump got it right from the start when he issued an executive order in January to strengthen America's AI – the next great technological forefront. From Day One as Environmental Protection Agency (EPA) administrator, it was clear that EPA would have a major hand in permitting reform to cut down barriers that have acted as a roadblock so we can bolster the growth of AI and make America the AI capital of the world. In fact, it's an endeavor so important, it is a core pillar of my Powering the Great American Comeback initiative.