In the face of rapidly advancing AI technology, individuals will increasingly rely on AI agents to navigate life's growing complexities, raising critical concerns about maintaining both human agency and autonomy. This paper addresses a fundamental dilemma posed by AI decision-support systems: the risk of either becoming overwhelmed by complex decisions, thus losing agency, or having autonomy compromised by externally controlled choice architectures reminiscent of ``nudging'' practices. While the ``nudge'' framework, based on the use of choice-framing to guide individuals toward presumed beneficial outcomes, initially appeared to preserve liberty, at AI-driven scale, it threatens to erode autonomy. To counteract this risk, the paper proposes a philosophic turn in AI design. AI should be constructed to facilitate decentralized truth-seeking and open-ended inquiry, mirroring the Socratic method of philosophical dialogue. By promoting individual and collective adaptive learning, such AI systems would empower users to maintain control over their judgments, augmenting their agency without undermining autonomy. The paper concludes by outlining essential features for autonomy-preserving AI systems, sketching a path toward AI systems that enhance human judgment rather than undermine it.

Background: Emergency department (ED) overcrowding remains a major challenge, causing delays in care and increased operational strain. Hospital management often reacts to congestion after it occurs. Machine learning predictive modeling offers a proactive approach by forecasting patient flow metrics, such as waiting count, to improve resource planning and hospital efficiency. Objective: This study develops machine learning models to predict ED waiting room occupancy at two time scales. The hourly model forecasts the waiting count six hours ahead (e.g., a 1 PM prediction for 7 PM), while the daily model estimates the average waiting count for the next 24 hours (e.g., a 5 PM prediction for the following day's average). These tools support staffing decisions and enable earlier interventions to reduce overcrowding. Methods: Data from a partner hospital's ED in the southeastern United States were used, integrating internal metrics and external features. Eleven machine learning algorithms, including traditional and deep learning models, were trained and evaluated. Feature combinations were optimized, and performance was assessed across varying patient volumes and hours. Results: TSiTPlus achieved the best hourly prediction (MAE: 4.19, MSE: 29.32). The mean hourly waiting count was 18.11, with a standard deviation of 9.77. Accuracy varied by hour, with MAEs ranging from 2.45 (11 PM) to 5.45 (8 PM). Extreme case analysis at one, two, and three standard deviations above the mean showed MAEs of 6.16, 10.16, and 15.59, respectively. For daily predictions, XCMPlus performed best (MAE: 2.00, MSE: 6.64), with a daily mean of 18.11 and standard deviation of 4.51. Conclusions: These models accurately forecast ED waiting room occupancy and support proactive resource allocation. Their implementation has the potential to improve patient flow and reduce overcrowding in emergency care settings.

This paper presents two attack strategies designed to evade detection in ADMM-based systems by preventing significant changes to the residual during the attacked iteration. While many detection algorithms focus on identifying false data injection through residual changes, we show that our attacks remain undetected by keeping the residual largely unchanged. The first strategy uses a random starting point combined with Gram-Schmidt orthogonalization to ensure stealth, with potential for refinement by enhancing the orthogonal component to increase system disruption. The second strategy builds on the first, targeting financial gains by manipulating reactive power and pushing the system to its upper voltage limit, exploiting operational constraints. The effectiveness of the proposed attack-resilient mechanism is demonstrated through case studies on the IEEE 14-bus system. A comparison of the two strategies, along with commonly used naive attacks, reveals trade-offs between simplicity, detectability, and effectiveness, providing insights into ADMM system vulnerabilities. These findings underscore the need for more robust monitoring algorithms to protect against advanced attack strategies.

Deep learning models, especially large Transformers, carry substantial "memory" in their intermediate layers -- an \emph{internal world} that encodes a wealth of relational and contextual knowledge. This work harnesses that internal world for wildfire occurrence prediction by introducing a modular architecture built upon Gemma 3, a state-of-the-art multimodal model. Rather than relying on Gemma 3's original embedding and positional encoding stacks, we develop a custom feed-forward module that transforms tabular wildfire features into the hidden dimension required by Gemma 3's mid-layer Transformer blocks. We freeze these Gemma 3 sub-layers -- thus preserving their pretrained representation power -- while training only the smaller input and output networks. This approach minimizes the number of trainable parameters and reduces the risk of overfitting on limited wildfire data, yet retains the benefits of Gemma 3's broad knowledge. Evaluations on a Moroccan wildfire dataset demonstrate improved predictive accuracy and robustness compared to standard feed-forward and convolutional baselines. Ablation studies confirm that the frozen Transformer layers consistently contribute to better representations, underscoring the feasibility of reusing large-model mid-layers as a learned internal world. Our findings suggest that strategic modular reuse of pretrained Transformers can enable more data-efficient and interpretable solutions for critical environmental applications such as wildfire risk management.

Large Language Models (LLMs) have exhibited impressive natural language processing capabilities but often perpetuate social biases inherent in their training data. To address this, we introduce MultiLingual Augmented Bias Testing (MLA-BiTe), a framework that improves prior bias evaluation methods by enabling systematic multilingual bias testing. MLA-BiTe leverages automated translation and paraphrasing techniques to support comprehensive assessments across diverse linguistic settings. In this study, we evaluate the effectiveness of MLA-BiTe by testing four state-of-the-art LLMs in six languages -- including two low-resource languages -- focusing on seven sensitive categories of discrimination.

--This work presents a novel method for leveraging intercepted Radio Frequency (RF) signals to inform a constrained Reinforcement Learning (RL) policy for robust control of a satellite operating in contested environments. Uncooperative satellite engagements with nation-state actors prompts the need for enhanced maneuverability and agility on-orbit. However, robust, autonomous and rapid adversary avoidance capabilities for the space environment is seldom studied. Further, the capability constrained nature of many space vehicles does not afford robust space situational awareness capabilities that can be used for well informed maneuvering. We present a "Cat & Mouse" system for training optimal adversary avoidance algorithms using RL. We propose the novel approach of utilizing intercepted radio frequency communication and dynamic spacecraft state as multi-modal input that could inform paths for a mouse to outmaneuver the cat satellite. Given the current ubiquitous use of RF communications, our proposed system can be applicable to a diverse array of satellites. In addition to providing a comprehensive framework for training and implementing a constrained RL policy capable of providing control for robust adversary avoidance, we also explore several optimization based methods for adversarial avoidance. These methods were then tested on real-world data obtained from the Space Surveillance Network (SSN) to analyze the benefits and limitations of different avoidance methods. In March of 2025, Chinese satellites exhibited dog-fighting capabilities [1], following years of both Russian [2] and Chinese [3] satellites approaching dangerously close to US satellites in geosynchronous orbit. Such uncooperative activity prompts the need for satellite agility and maneuverability which can be facilitated through edge-based autonomy. To achieve this, appropriate sensing would be required to properly characterize the contested environment. Not all satellites have precise space domain awareness (SDA) sensing suites onboard, despite having powerful buses and flight controllers that can facilitate autonomous operations. We propose leveraging an uncooperative space vehicle's communication systems as a means to evaluate safe flight control policies to carefully navigate contested domains in situations where support from the ground is not feasible.

Elon Musk and his companies face at least 2.37bn in legal exposure from federal investigations, litigation and regulatory oversight, according to a new report from Senate Democrats. The report attempts to put a number to Musk's many conflicts of interest through his work with his so-called "department of government efficiency" (Doge), warning that he may seek to use his influence to avoid legal liability. The report, which was published on Monday by Democratic members of the Senate homeland security committee's permanent subcommittee on investigations, looked at 65 actual or potential actions against Musk across 11 separate agencies. Investigators calculated the financial liabilities Musk and his companies, such as Tesla, SpaceX and Neuralink, may face in 45 of those actions. Since Donald Trump won re-election last year and Musk took on the role of de facto head of Doge in January, ethics watchdogs and Democratic officials have warned that the Tesla CEO could use his power to oust regulators and quash investigations into his companies.

Thousands of civil servants at the heart of the UK government, including those working directly to support Prime Minister Keir Starmer, are using a proprietary artificial intelligence chatbot to carry out their work, New Scientist can reveal. Officials have refused to disclose on the record exactly how the tool is being used, whether the prime minister is receiving advice that has been prepared using AI or how civil servants are mitigating the risks of inaccurate or biased AI outputs. Experts say the lack of disclosure raises concerns about government transparency and the accuracy of information being used in government. After securing the world-first release of ChatGPT logs under freedom of information (FOI) legislation, New Scientist asked 20 government departments for records of their interactions with Redbox, a generative AI tool developed in house and trialled among UK government staff. The large language model-powered chatbot allows users to interrogate government documents and to "generate first drafts of briefings", according to one of the people behind its development.

The UK's Children's Commissioner is calling for a ban on AI deepfake apps that create nude or sexual images of children, according to a new report. It states that such "nudification" apps have become so prevalent that many girls have stopped posting photos on social media. And though creating or uploading CSAM images is illegal, apps used to create deepfake nude images are still legal. "Children have told me they are frightened by the very idea of this technology even being available, let alone used. They fear that anyone -- a stranger, a classmate, or even a friend -- could use a smartphone as a way of manipulating them by creating a naked image using these bespoke apps." said Children's Commissioner Dame Rachel de Souza.

The Trump administration has laid out its own ambitious goals for recruiting more tech talent. On April 3, Russell Vought, Trump's Director of the Office of Management and Budget, released a 25-page memo for how federal leaders were expected to accelerate the government's use of AI. "Agencies should focus recruitment efforts on individuals that have demonstrated operational experience in designing, deploying, and scaling AI systems in high-impact environments," Vought wrote. Putting that into action will be harder than it needed to be, says Deirdre Mulligan, who directed the National Artificial Intelligence Initiative Office in the Biden White House. "The Trump Administration's actions have not only denuded the government of talent now, but I'm sure that for many folks, they will think twice about whether or not they want to work in government," Mulligan says. "It's really important to have stability, to have people's expertise be treated with the level of respect it ought to be and to have people not be wondering from one day to the next whether they're going to be employed."