The UK AI Safety Institute (UK AISI) and its parallel organisation in the United States (US AISI) take up a unique position in the recently established International Network of AISIs. Both are in jurisdictions with frontier AI companies and are assuming leading roles in the international conversation on AI Safety. This paper argues that it is in the interest of both institutions to share specific categories of information with the International Network of AISIs, deliberately abstain from sharing others and carefully evaluate sharing some categories on a case by case basis, according to domestic priorities. The paper further proposes a provisional framework with which policymakers and researchers can distinguish between these three cases, taking into account the potential benefits and risks of sharing specific categories of information, ranging from pre-deployment evaluation results to evaluation standards. In an effort to further improve the research on AI policy relevant information sharing decisions, the paper emphasises the importance of continuously monitoring fluctuating factors influencing sharing decisions and a more in-depth analysis of specific policy relevant information
This technical report presents methods developed by the UK AI Security Institute for assessing whether advanced AI systems reliably follow intended goals. Specifically, we evaluate whether frontier models sabotage safety research when deployed as coding assistants within an AI lab. Applying our methods to four frontier models, we find no confirmed instances of research sabotage. However, we observe that Claude Opus 4.5 Preview (a pre-release snapshot of Opus 4.5) and Sonnet 4.5 frequently refuse to engage with safety-relevant research tasks, citing concerns about research direction, involvement in self-training, and research scope. We additionally find that Opus 4.5 Preview shows reduced unprompted evaluation awareness compared to Sonnet 4.5, while both models can distinguish evaluation from deployment scenarios when prompted. Our evaluation framework builds on Petri, an open-source LLM auditing tool, with a custom scaffold designed to simulate realistic internal deployment of a coding agent. We validate that this scaffold produces trajectories that all tested models fail to reliably distinguish from real deployment data. We test models across scenarios varying in research motivati
Artificial Intelligence (AI) Safety Institutes and governments worldwide are deciding whether they evaluate advanced AI themselves, support a private evaluation ecosystem or do both. Evaluation regimes have been established in a wide range of industry contexts to monitor and evaluate firms' compliance with regulation. Evaluation is a necessary governance tool to understand and manage the risks of a technology. This paper draws from nine such regimes to inform (i) who should evaluate which parts of advanced AI; and (ii) how much capacity public bodies may need to evaluate advanced AI effectively. First, the effective responsibility distribution between public and private evaluators depends heavily on specific industry and evaluation conditions. On the basis of advanced AI's risk profile, the sensitivity of information involved in the evaluation process, and the high costs of verifying safety and benefit claims of AI Labs, we recommend that public bodies become directly involved in safety critical, especially gray- and white-box, AI model evaluations. Governance and security audits, which are well-established in other industry contexts, as well as black-box model evaluations, may be
In recent times Mechanical and Production industries are facing increasing challenges related to the shift toward sustainable manufacturing. In this article, machining was performed in dry cutting condition with a newly developed coated insert called AlTiSiN coated carbides coated through scalable pulsed power plasma technique in dry cutting condition and a dataset was generated for different machining parameters and output responses. The machining parameters are speed, feed, depth of cut and the output responses are surface roughness, cutting force, crater wear length, crater wear width, and flank wear. The data collected from the machining operation is used for the development of machine learning (ML) based surrogate models to test, evaluate and optimize various input machining parameters. Different ML approaches such as polynomial regression (PR), random forest (RF) regression, gradient boosted (GB) trees, and adaptive boosting (AB) based regression are used to model different output responses in the hard machining of AISI D6 steel. The surrogate models for different output responses are used to prepare a complex objective function for the germinal center algorithm-based optimiz
In November 2023, the UK and US announced the creation of their AI Safety Institutes (AISIs). Five other jurisdictions have followed in establishing AISIs or similar institutions, with more likely to follow. While there is considerable variation between these institutions, there are also key similarities worth identifying. This primer describes one cluster of similar AISIs, the "first wave," consisting of the Japan, UK, and US AISIs. First-wave AISIs have several fundamental characteristics in common: they are technical government institutions, have a clear mandate related to the safety of advanced AI systems, and lack regulatory powers. Safety evaluations are at the center of first-wave AISIs. These techniques test AI systems across tasks to understand their behavior and capabilities on relevant risks, such as cyber, chemical, and biological misuse. They also share three core functions: research, standards, and cooperation. These functions are critical to AISIs' work on safety evaluations but also support other activities such as scientific consensus-building and foundational AI safety research. Despite its growing popularity as an institutional model, the AISI model is not free f
The rapid rise of autonomous AI systems and advancements in agent capabilities are introducing new risks due to reduced oversight of real-world interactions. Yet agent testing remains nascent and is still a developing science. As AI agents begin to be deployed globally, it is important that they handle different languages and cultures accurately and securely. To address this, participants from The International Network for Advanced AI Measurement, Evaluation and Science, including representatives from Singapore, Japan, Australia, Canada, the European Commission, France, Kenya, South Korea, and the United Kingdom have come together to align approaches to agentic evaluations. This is the third exercise, building on insights from two earlier joint testing exercises conducted by the Network in November 2024 and February 2025. The objective is to further refine best practices for testing advanced AI systems. The exercise was split into two strands: (1) common risks, including leakage of sensitive information and fraud, led by Singapore AISI; and (2) cybersecurity, led by UK AISI. A mix of open and closed-weight models were evaluated against tasks from various public agentic benchmarks.
Surface wettability, governed by both topography and chemistry, plays a critical role in applications such as heat transfer, lubrication, microfluidics, and surface coatings. In this study, we present a machine learning (ML) framework capable of accurately predicting the wettability of laser-textured metal alloys using experimentally derived morphological and chemical features. Superhydrophilic and superhydrophobic surfaces were fabricated on AA6061 and AISI 4130 alloys via nanosecond laser texturing followed by chemical immersion treatments. Surface morphology was quantified using the Laws texture energy method and profilometry, while surface chemistry was characterized through X-ray photoelectron spectroscopy (XPS), extracting features such as functional group polarity, molecular volume, and peak area fraction. These features were used to train an ensemble neural network model incorporating residual connections, batch normalization, and dropout regularization. The model achieved high predictive accuracy (R2 = 0.942, RMSE = 13.896), outperforming previous approaches. Feature importance analysis revealed that surface chemistry had the strongest influence on contact angle prediction
As frontier AI models are deployed globally, it is essential that their behaviour remains safe and reliable across diverse linguistic and cultural contexts. To examine how current model safeguards hold up in such settings, participants from the International Network for Advanced AI Measurement, Evaluation and Science, including representatives from Singapore, Japan, Australia, Canada, the EU, France, Kenya, South Korea and the UK conducted a joint multilingual evaluation exercise. Led by Singapore AISI, two open-weight models were tested across ten languages spanning high and low resourced groups: Cantonese English, Farsi, French, Japanese, Korean, Kiswahili, Malay, Mandarin Chinese and Telugu. Over 6,000 newly translated prompts were evaluated across five harm categories (privacy, non-violent crime, violent crime, intellectual property and jailbreak robustness), using both LLM-as-a-judge and human annotation. The exercise shows how safety behaviours can vary across languages. These include differences in safeguard robustness across languages and harm types and variation in evaluator reliability (LLM-as-judge vs. human review). Further, it also generated methodological insights for
Large language models (LLMs) are increasingly used to simulate human behavior in social settings such as legal mediation, negotiation, and dispute resolution. However, it remains unclear whether these simulations reproduce the personality-behavior patterns observed in humans. Human personality, for instance, shapes how individuals navigate social interactions, including strategic choices and behaviors in emotionally charged interactions. This raises the question: Can LLMs, when prompted with personality traits, reproduce personality-driven differences in human conflict behavior? To explore this, we introduce an evaluation framework that enables direct comparison of human-human and LLM-LLM behaviors in dispute resolution dialogues with respect to Big Five Inventory (BFI) personality traits. This framework provides a set of interpretable metrics related to strategic behavior and conflict outcomes. We additionally contribute a novel dataset creation methodology for LLM dispute resolution dialogues with matched scenarios and personality traits with respect to human conversations. Finally, we demonstrate the use of our evaluation framework with three contemporary closed-source LLMs and
A gas-driven permeation (GDP) platform, SHIELD (Salt-compatible Hydrogen barrier Investigation and EvaLuation for fusion Devices), has been developed to measure hydrogen transport properties in structural materials under controlled thermal and pressure conditions. The system is designed to minimise experimental uncertainties associated with leaks, temperature instability, and pressure measurement, while providing reproducible conditions for permeation experiments. The rig operates in a static GDP configuration with independent upstream and downstream volumes, enabling precise control of driving pressure and accurate measurement of downstream pressure rise. An openly documented data acquisition and processing framework is implemented to ensure data traceability and reproducibility. The platform's performance is demonstrated by hydrogen permeation measurements on 316 stainless steel and AISI 1018 low-carbon steel over the temperature range of \SIrange{100}{600}{\degreeCelsius}. Steady-state permeation fluxes are extracted from linear downstream pressure rise and used to determine permeability. The measured permeability exhibits Arrhenius behaviour and agrees well with published liter
Surface roughness plays a critical role in ultrashort pulse laser ablation, particularly for industrial applications using burst mode operations, multi-pulse laser processing, and the generation of laser-induced periodic surface structures. Hence, we address the impact of surface roughness on the resulting laser ablation topography predicted by a simulation model and compared to experimental results. We present a comprehensive multi-scale simulation framework that first employs finite-difference-time-domain simulations for calculating the surface fluence distribution on a rough surface measured by an atomic-force-microscope followed by the two-temperature model coupled with hydrodynamic/solid mechanics simulation for the initial material heating. Lastly, a computational fluid dynamics model for material relaxation and fluid flow is developed and employed. Final state results of aluminum and AISI 304 stainless steel simulations demonstrated alignment with established ablation models and crater dimension prediction. Notably, Al exhibited significant optical scattering effects due to initial surface roughness of 15 nm - being 70 times below the laser wavelength, leading to localized,
Next-generation gravitational wave detectors (GWDs) such as the Cosmic Explorer and Einstein Telescope demand extensive ultra-high vacuum systems, making material cost and performance critical considerations. This study investigates the potential of ferritic stainless steel as a cost-effective alternative to the commonly used austenitic stainless steel for UHV components, focusing on the analysis of outgassing rates pre and post-bakeout at 80°C and 150°C for 48 hours. The tested ferritic stainless steels exhibit significantly lower hydrogen content than standard AISI 304L steel. After bakeout, the hydrogen outgassing rates - measured down to 10$^{-15}$ mbar l s$^{-1}$ cm$^{-2}$ - are three orders of magnitude lower than those of similarly conditioned austenitic stainless steels. These results highlight ferritic stainless steel as a promising, economical, and high-performance candidate for future GWDs vacuum systems.
International standards are crucial for ensuring that frontier AI systems are developed and deployed safely around the world. Since the AI Safety Institutes (AISIs) possess in-house technical expertise, mandate for international engagement, and convening power in the national AI ecosystem while being a government institution, we argue that they are particularly well-positioned to contribute to the international standard-setting processes for AI safety. In this paper, we propose and evaluate three models for AISI involvement: 1. Seoul Declaration Signatories, 2. US (and other Seoul Declaration Signatories) and China, and 3. Globally Inclusive. Leveraging their diverse strengths, these models are not mutually exclusive. Rather, they offer a multi-track system solution in which the central role of AISIs guarantees coherence among the different tracks and consistency in their AI safety focus.
Surface topography governs tribological performance, yet conventional parameters describe either amplitude statistics or spectral content in isolation. We introduce a scale-dependent framework that represents surface height and directional gradient as conjugate coordinates of a structural phase space. The elastic reference energy derived from Persson's contact mechanics theory defines a metric that couples surface geometry to the elastic half-space response. A maximum-entropy formulation yields a canonical state density. In the Gaussian limit this formulation recovers Persson's spectral description exactly, showing that the power spectral density is a complete contact mechanical descriptor only under Gaussian statistics. The associated Lagrange multiplier defines a topographic temperature in the sense of Grmela's multiscale thermodynamics and embeds the areal subsystem within a scale-dependent boundary potential. The framework is validated experimentally using ground and honed AISI 52100 steel discs before and after running in. The ground surface contracts toward lower entropy and elastic energy, whereas the honed surface expands into previously unoccupied states. These opposite tr
Image classification systems often inherit biases from uneven group representation in training data. For example, in face datasets for hair color classification, blond hair may be disproportionately associated with females, reinforcing stereotypes. A recent approach leverages the Stable Diffusion model to generate balanced training data, but these models often struggle to preserve the original data distribution. In this work, we explore multiple diffusion-finetuning techniques, e.g., LoRA and DreamBooth, to generate images that more accurately represent each training group by learning directly from their samples. Additionally, in order to prevent a single DreamBooth model from being overwhelmed by excessive intra-group variations, we explore a technique of clustering images within each group and train a DreamBooth model per cluster. These models are then used to generate group-balanced data for pretraining, followed by fine-tuning on real data. Experiments on multiple benchmarks demonstrate that the studied finetuning approaches outperform vanilla Stable Diffusion on average and achieve results comparable to SOTA debiasing techniques like Group-DRO, while surpassing them as the dat
As part of nationwide efforts aligned with the United Nations' Sustainable Development Goal 3 on Universal Health Coverage, Ethiopia's Ministry of Health is strengthening health posts to expand access to essential healthcare services. However, only a fraction of this health system strengthening effort can be implemented each year due to limited budgets and other competing priorities, thus the need for an optimization framework to guide prioritization across the regions of Ethiopia. In this paper, we develop a tool, Health Access Resource Planner (HARP), based on a principled decision-support optimization framework for sequential facility planning that aims to maximize population coverage under budget uncertainty while satisfying region-specific proportionality targets at every time step. We then propose two algorithms: (i) a learning-augmented approach that improves upon expert recommendations at any single-step; and (ii) a greedy algorithm for multi-step planning, both with strong worst-case approximation estimation. In collaboration with the Ethiopian Public Health Institute and Ministry of Health, we demonstrated the empirical efficacy of our method on three regions across vario
Antibiotic Resistance (AR) is a critical global health challenge that necessitates the development of cost-effective, efficient, and accurate diagnostic tools. Given the genetic basis of AR, techniques such as Polymerase Chain Reaction (PCR) that target specific resistance genes offer a promising approach for predictive diagnostics using a limited set of key genes. This study introduces GenoARM, a novel framework that integrates reinforcement learning (RL) with transformer-based models to optimize the selection of PCR gene tests and improve AR predictions, leveraging observed metadata for improved accuracy. In our evaluation, we developed several high-performing baselines and compared them using publicly available datasets derived from real-world bacterial samples representing multiple clinically relevant pathogens. The results show that all evaluated methods achieve strong and reliable performance when metadata is not utilized. When metadata is introduced and the number of selected genes increases, GenoARM demonstrates superior performance due to its capacity to approximate rewards for unseen and sparse combinations. Overall, our framework represents a major advancement in optimiz
Sea Surface Temperature (SST) is crucial for understanding upper-ocean thermal dynamics and ocean-atmosphere interactions, which have profound economic and social impacts. While data-driven models show promise in SST prediction, their black-box nature often limits interpretability and overlooks key physical processes. Recently, physics-informed neural networks have been gaining momentum but struggle with complex ocean-atmosphere dynamics due to 1) inadequate characterization of seawater movement (e.g., coastal upwelling) and 2) insufficient integration of external SST drivers (e.g., turbulent heat fluxes). To address these challenges, we propose SSTODE, a physics-informed Neural Ordinary Differential Equations (Neural ODEs) framework for SST prediction. First, we derive ODEs from fluid transport principles, incorporating both advection and diffusion to model ocean spatiotemporal dynamics. Through variational optimization, we recover a latent velocity field that explicitly governs the temporal dynamics of SST. Building upon ODE, we introduce an Energy Exchanges Integrator (EEI)-inspired by ocean heat budget equations-to account for external forcing factors. Thus, the variations in t
Ensuring web accessibility is crucial for advancing social welfare, justice, and equality in digital spaces, yet the vast majority of website user interfaces remain non-compliant, due in part to the resource-intensive and unscalable nature of current auditing practices. While WCAG-EM offers a structured methodology for site-wise conformance evaluation, it involves great human efforts and lacks practical support for execution at scale. In this work, we present an auditing framework, AAA, which operationalizes WCAG-EM through a human-AI partnership model. AAA is anchored by two key innovations: GRASP, a graph-based multimodal sampling method that ensures representative page coverage via learned embeddings of visual, textual, and relational cues; and MaC, a multimodal large language model-based copilot that supports auditors through cross-modal reasoning and intelligent assistance in high-effort tasks. Together, these components enable scalable, end-to-end web accessibility auditing, empowering human auditors with AI-enhanced assistance for real-world impact. We further contribute four novel datasets designed for benchmarking core stages of the audit pipeline. Extensive experiments de
Illicit Massage Businesses (IMBs) are a covert and persistent form of organized exploitation that operate under the facade of legitimate wellness services while facilitating human trafficking, sexual exploitation, and coerced labor. Detecting IMBs is difficult due to encoded digital advertisements, frequent changes in personnel and locations, and the reuse of shared infrastructure such as phone numbers and addresses. Traditional approaches, including community tips and regulatory inspections, are largely reactive and ineffective at revealing the broader operational networks traffickers rely on. To address these challenges, we introduce IMBWatch, a spatio-temporal graph neural network (ST-GNN) framework for large-scale IMB detection. IMBWatch constructs dynamic graphs from open-source intelligence, including scraped online advertisements, business license records, and crowdsourced reviews. Nodes represent heterogeneous entities such as businesses, aliases, phone numbers, and locations, while edges capture spatio-temporal and relational patterns, including co-location, repeated phone usage, and synchronized advertising. The framework combines graph convolutional operations with tempo