arXiv:2606.02859v1 Announce Type: new Abstract: How can a population of agents self-orchestrate and self-adapt into stronger collective intelligence without centralized control? Inspired by Friedrich Hayek's economic theory of decentralized coordination in markets, we study this question through an agent economy in which agents compete via auctions for the right to act, exchange payments, and accumulate wealth from environmental rewards. These simple economic signals induce decentralized credit assignment, driving planning without global orchestration or explicit communication protocols. The population evolves through economic selection: effective agents accumulate wealth and are mutated via exploitation, while ineffective ones go bankrupt and are replaced via exploration. We show that, initialized with weak agents, the economy produces emergent multi-step reasoning strategies and outperforms stronger monolithic baselines across five agentic tasks, including mathematical reasoning, financial research, scientific research, accelerator design, and distributed-system optimization. We further provide theoretical insights into how economic dynamics shape agent behaviors, linking local incentives to long-term global performance. Our results suggest a new path to multi-agent intelligence: rather than engineering coordination, we can design decentralized incentive structures under which it automatically emerges.

arXiv:2606.02597v1 Announce Type: new Abstract: The development of brain-computer interfaces (BCIs) based on electroencephalograms (EEGs) has advanced significantly mainly to machine learning. Although the majority of earlier research has been on increasing classification accuracy, relatively little focus has been placed on security and robustness. According to recent research, EEG-based BCIs are susceptible to adversarial attacks, which can cause misdiagnosis due to minute, well-crafted disturbances. Evaluating model robustness against such perturbations is therefore critical for ensuring reliable deployment. In this study, we propose a lightweight custom Convolutional Neural Network (CNN) architecture to investigate adversarial robustness in EEG-based BCIs. The suggested method is assessed using two EEG datasets and contrasted with three novel CNN models tailored to EEG, namely EEGNet, DeepConvNet, and SleepEEGNet, under gradient-based adversarial attack scenarios. According to experimental findings, the suggested model continuously performs better in classification under adversarial perturbations compared to baseline models, indicating improved robustness. These findings highlight the potential of lightweight architectures for enhancing the reliability of EEG-based BCI systems under adversarial conditions.

arXiv:2606.02764v1 Announce Type: new Abstract: Satellite-derived bathymetry (SDB) from multispectral imagery is cost-effective but scales poorly across regions, especially in optically complex coastal environments. We evaluate machine learning and deep learning for transferable SDB over the 0-20 m depth range using Sentinel-2 imagery. A Random Forest baseline and four CNNs (ResNet-50, ResNet-101, EfficientNet-B4, ConvNeXt-Large) are trained on Pratas Island and selected Great Barrier Reef regions, then evaluated on spatially independent intra- and cross-regional test areas. Preserving spatial continuity during training, by keeping contiguous reef blocks rather than random patches, is the single most impactful design choice; we further introduce a Smooth Weight Function (SWF)-weighted RMSE loss that emphasizes near-surface depths. With these choices, intra-regional RMSE ranges from 1.15 to 1.92 m over 0-20 m and is as low as 0.26 m for depths 2.99-3.78 m), while the deep models stay more robust (2.46-2.98 m). On the public MagicBathyNet aerial-RGB benchmark (0-16 m) the proposed networks reach 0.19-0.22 m RMSE, outperforming a U-Net baseline and a task-specific transformer architecture with substantially fewer parameters. We further exploit multi-temporal repeat imagery: training on it broadens diversity, and median-aggregating predictions across passes at inference reduces noise from changing sun angles, atmospheric conditions, water properties, and tides. We release optimized architectures and pretrained weights to enable scalable transfer to new sites.

arXiv:2606.02747v1 Announce Type: new Abstract: Planning records define restrictions over geographic areas, but their source documents often provide only indirect spatial evidence rather than machine-readable boundaries. We introduce Plan2Map, a 208-case multimodal benchmark for document-grounded geospatial boundary reconstruction from UK planning records. Given only a source planning document, systems must reconstruct a valid geospatial boundary from notice text, schedules, map plates, map labels, and boundary annotations; the reference GeoJSON is held out for scoring. We propose GeoPlanAgent, a document-grounded, geospatial-tool-in-the-loop system that decomposes the task into evidence extraction, localisation, map registration, boundary segmentation, projection, and verification. On Plan2Map, GeoPlanAgent achieves 0.736 mean IoU and 0.904 median IoU, with 67.8\% of predictions at or above 0.8 IoU, substantially outperforming direct VLM-to-GeoJSON baselines. Diagnostic analysis shows that direct VLM prediction remains unreliable, while remaining errors are concentrated in localisation and map registration, and supervised boundary segmentation substantially improves pixel-level mask quality. Plan2Map provides a concrete testbed for multimodal geospatial reconstruction from public planning records. Project page: https://odeb1.github.io/Plan2Map_Project_Page/.

arXiv:2606.02753v1 Announce Type: new Abstract: Video world models are a foundational generative technology for embodied AI and the Metaverse, yet existing approaches are inherently limited to a single agent observing from a single perspective. Extending these models to multi-agent settings introduces two critical challenges: data scarcity (coordinated multi-view recordings are prohibitively expensive to collect for general open-domain scenarios) and world state alignment (independently generated video streams cannot ensure that shared physical environments and events evolve consistently across views). To address these challenges, we propose MetaWorld, a novel framework that scales multi-agent video world models to open-domain environments directly from single-view videos. First, we introduce Monocular World-State Unrolling (MWSU) to explicitly decompose monocular footage into the camera operator's ego-motion and the visible subject's spatial trajectory. This camera-trajectory decomposition naturally extracts synchronized multi-agent motion data within a shared 3D space, completely bypassing the need for multi-camera setups. Second, for precise visual control, we develop the Subject-Aware World Generator to enable appearance-driven simulation conditioned on per-agent identity images. Finally, to ensure both views are grounded in the identical physical reality, we propose World-State Alignment, a per-frame inter-branch cross-attention mechanism inserted at every transformer layer of the video DiT. By jointly synchronizing the denoising process, WSA enforces both static geometric consistency and dynamic motion consistency, encouraging that the shared 3D environment and physical events remain well-aligned across both egocentric views. Extensive experiments demonstrate that MetaWorld achieves superior cross-view consistency and identity fidelity, establishing a highly scalable, physics-driven paradigm for multi-agent video world modeling.

arXiv:2606.02598v1 Announce Type: new Abstract: Accurate and generalizable estimation of cognitive workload from electroencephalography (EEG) is critical for human-centered and safety-critical systems. Although EEG is widely used for workload assessment, the consistency of region-level EEG contributions across tasks, datasets, and subjects remains unclear. This paper presents a region-level evaluation framework for EEG-based workload prediction in which models are trained and evaluated using features extracted exclusively from electrodes belonging to anatomically defined scalp regions. We perform a large-scale analysis across four publicly available EEG workload datasets spanning diverse task demands, recording hardware, and electrode montages. Region importance is quantified using a model-agnostic, performance-based approach under both mixed-subject and subject-independent evaluation protocols, with results aggregated using a rank-based strategy to ensure robustness across experimental configurations. Across all datasets and subject-independent evaluations, frontal electrode groups outperform the full-scalp baseline by approximately 15-20% in relative rank position while using substantially fewer electrodes. Fronto-central regions exhibit the most stable predictive utility, whereas posterior and occipital regions contribute less consistently across experimental conditions. These findings indicate that workload-relevant EEG information is most consistently retained within frontal and fronto-central electrode groups, supporting the design of efficient and generalizable EEG-based workload monitoring systems.

arXiv:2606.02603v1 Announce Type: new Abstract: Camouflaged object detection has improved substantially, but most standard benchmarks evaluate models only on clean images. This is not realistic because real cameras often capture blur, sensor noise, weather effects, and compression artifacts. We present COD10K-C, a corruption robustness benchmark based on COD10K. It includes 8 corruption types and 5 severity levels, giving 40 conditions and 81,040 evaluation pairs in total. We evaluate three popular camouflaged object detection models, SINet-v2, PFNet, and ZoomNet, as well as a lightweight model called RobustCODLite. All models show clear performance drops on corrupted images. Motion blur and Gaussian blur cause the largest drops, with SINet-v2 losing 18.5 Dice points under motion blur. Brightness and fog are less harmful. RobustCODLite uses corruption augmentation, a frequency-prior branch, and an uncertainty-consistency loss. It retains 92.3% of its clean Dice score under corruption, compared with 87.7% for SINet-v2, 84.8% for ZoomNet, and 84.1% for PFNet. On the hardest corruptions, RobustCODLite matches or outperforms models that perform better on clean data. We will release the COD10K-C GitHub repository to support future research in robust camouflaged object detection.

arXiv:2606.02584v1 Announce Type: new Abstract: Idiomatic expressions remain a persistent challenge for natural language processing because their meanings are often non-compositional, context-dependent, and difficult to align across languages. Existing idiom resources are often limited in scale, contextual diversity, or multilingual coverage, restricting their utility for modern language models. We introduce IdiomX, a large-scale multilingual benchmark for idiom understanding, retrieval, and interpretation, constructed through a reproducible multi-stage pipeline combining lexical resource extraction, large-scale normalization, controlled large language model enrichment, and structured validation. The resulting dataset contains over 190K contextualized examples spanning 12K+ idioms, with aligned English, Arabic, and French semantic representations, idiomatic and literal usage labels, and rich linguistic metadata. Building on this resource, we define a unified four-task benchmark covering idiom detection, context-to-idiom retrieval, Arabic-to-English idiom retrieval, and idiom interpretation, extending evaluation from figurative recognition to semantic grounding and explainable meaning retrieval. Experiments show that contextual transformer models substantially improve idiom detection, while hybrid retrieval and reranking architectures significantly strengthen both monolingual and cross-lingual idiom retrieval. Results further demonstrate that idiom interpretation can be effectively modeled as a semantic retrieval task, introducing interpretability as a complementary benchmark dimension. Overall, IdiomX provides a scalable benchmark for studying idiomatic language as a progression from detection to retrieval and semantic interpretation, and offers a modular framework extensible to additional languages and figurative reasoning tasks

arXiv:2606.02601v1 Announce Type: new Abstract: Within-dataset class-split evaluation is widely used as a proxy for fully unconditional out-of-distribution anomaly detection. We show that this protocol can become ill-posed when the held-out anomaly class overlaps the normal mixture in representation space. In this regime, anomaly scores may collapse toward chance or even invert, and the preferred score direction can depend on the unknown anomaly class. We introduce a simple training-free diagnostic, neighborhood class leakage, and show that it predicts score-direction instability across Fashion-MNIST, CIFAR-10, and Imagenette, in both pixel and VAE latent spaces. Our results suggest that class-split AD benchmarks should be treated as geometry-dependent stress tests rather than unconditional evidence of anomaly-detection ability.

arXiv:2606.02789v1 Announce Type: new Abstract: Human-object interaction (HOI) recognition is critical for automatically analyzing student behavior in complex educational environments. Although state-of-the-art (SOTA) HOI detectors perform well on benchmark datasets, their performance often degrades when deployed in real-world training environments due to domain-specific objects, occlusions, and complex visual conditions. In this paper, we introduce a diagnosis-driven framework that integrates a triplet-level HOI error taxonomy with error-factor attribution analysis for real-world educational video data. We study this problem in the context of Critical Care Air Transport Team (CCATT) mixed-reality medical training. Based on an analysis of HOI failure modes and their causes, we develop a diagnosis-informed refinement strategy for adapting pretrained HOI models to the target domain. Experiments on the CCATT dataset show that this approach improves the macro-F1 score of a pretrained CDN model from 48.6 to 90.2 through targeted refinement guided by diagnosed error factors. These results highlight the value of detailed diagnostic analysis for informing targeted adaptation of HOI models in real-world educational environments.

arXiv:2606.02741v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used in sustainability-related decision support, reporting, and public communication, yet little systematic evidence exists on the environmental attitudes embedded in their outputs. This paper develops a benchmark for evaluating environmental cognition, affect, and behavioural recommendations in LLMs and applies it to 31 widely used proprietary and open-weight models. Drawing on questions from established environmental awareness surveys and additional sustainability-related behavioural measures, we compare LLM responses 1) among models and 2) between models and human survey benchmarks from Germany. We assess their robustness across prompting conditions. We find that many LLMs align more closely with environmentally progressive attitudes than the average survey respondent, exhibiting higher levels of environmental affect and cognition and recommending behaviours associated with substantial potential CO2 reductions. At the same time, we observe no systematic relationship between sustainability-oriented responses and model origin, size, or release context. However, models exhibit contextual sensitivity, controlled by persona-based prompting and show sycophantic shifts mirroring user-specified ideological positions, which raises concerns about steerability and normative reliability in real-world deployments. Our findings provide a reusable evaluation framework for assessing sustainability-related value alignment in LLMs and highlight the importance of governance, transparency, and critical oversight as AI systems become increasingly embedded in sustainability transformations and public decision-making.

arXiv:2606.02604v1 Announce Type: new Abstract: ESG and climate risk data remain fragmented across heterogeneous Scope 1, Scope 2, and Scope 3 reporting environments, while conventional validation pipelines lack provenance aware auditability, hidden drift detection, and reproducibility oriented governance. This paper proposes a deterministic climate risk intelligence framework integrating single source of truth orchestration, temporal anomaly detection, imbalance aware ensemble learning, and explainability oriented governance for auditable ESG validation. To support open reproducibility, we construct and release a synthetic ESG validation benchmark calibrated against publicly reported characteristics of the GHG Protocol, PCAF, and ISSB standards. The methodology incorporates temporal drift analysis, SMOTE based rare event optimization, ensemble learning, provenance aware orchestration, and TreeSHAP based interpretability for governance inspection and audit reconstruction. We evaluate the framework against statistical classifiers, anomaly detection methods, temporal forecasting baselines, and a threshold based system using classification metrics (recall, F1, ROC AUC), calibration metrics (ECE, Brier score), and a governance oriented audit trace completeness metric measuring the fraction of flagged anomalies for which a deterministic source to escalation provenance chain can be reconstructed. Results are reported as mean and standard deviation across stratified five fold cross validation with paired significance testing. The framework reframes ESG reporting toward deterministic climate risk governance infrastructure supporting reproducibility, explainability, and operational auditability.

arXiv:2606.02742v1 Announce Type: new Abstract: Spatial reasoning is fundamental to robotics, autonomy, and embodied AI, yet modern vision-language models (VLMs) remain unreliable on metric distance queries. A common assumption is that consistent predictions across viewpoints reflect geometric grounding. We test this assumption and find the opposite: leading VLMs often produce view-invariant and consistent answers even when those answers are incorrect, indicating weak coupling between predictions and viewpoint-specific visual evidence. We introduce \textbf{ViewDiag}, a controlled multi-view evaluation protocol built from Hypersim, ScanNet, and KITTI360, comprising 176 object-pair tracks across 80 scenes with 2--10 views per track. The protocol evaluates models along three axes: metric accuracy, distributional concentration, and a latent feature probe for internal collapse that distinguishes decision collapse from representation collapse. Across diverse models, we observe a consistent pattern of high prediction stability paired with substantial error, clustering in a regime characterized by strong consistency but low accuracy. \noindent These results challenge the common use of cross-view consistency as a proxy for geometric understanding. Instead, we show that stable predictions may reflect prior-driven collapse rather than evidence-sensitive reasoning. ViewDiag provides a controlled benchmark and diagnostic framework for evaluating spatial VLMs beyond accuracy alone. The code and data can be found \href{https://github.com/SDivakarBhat/Consistent_Yet_Wrong.git}{here}

arXiv:2606.02602v1 Announce Type: new Abstract: In computational pathology, Whole Slide Images (WSIs) survival analysis is crucial for patient prognosis assessment, but it faces multiple technical challenges. Although the Transformer captures long-range dependencies through its self-attention mechanism, its $O(N^2)$ time complexity causes a severe computational bottleneck in large-scale WSIs graph structures. The Mamba model breaks through the Transformer's computational bottleneck with linear complexity. But, owing to Mamba's high sensitivity to the order of input data, traditional node sorting methods in Graph Mamba, such as those based on node degree or subgraph size, fail to adequately account for the topological connectivity of graph data. This inadequacy consequently restricts the performance of Mamba's sequential modeling. Moreover, its unidirectional architecture cannot leverage the bidirectional spatial structure of images. To address these challenges, this paper proposes a novel Graph Mamba survival analysis framework based on topology-aware ordering (TopoMamSurv) to adapt to the sequential sensitivity of Mamba. Our visualization experiments further confirmed that the nodes extracted through the topology-aware ordering (TAO) strategy indeed exhibit higher similarity. Furthermore, we designed a bidirectional Mamba module and integrated a Graph Convolutional Network (GCN) to achieve bidirectional spatial context modeling of images, forming a hierarchical feature learning architecture for "local aggregation - global capture." This framework effectively reconciles the contradiction between long-range dependency modeling, computational efficiency, and spatial structure utilization in WSIs analysis through its systematic design of TAO, bidirectional semantic modeling, and hierarchical feature fusion. This framework has been validated for its comprehensive performance advantage on five TCGA datasets.

arXiv:2606.02750v1 Announce Type: new Abstract: Representations extracted from large language models (LLMs) play an important role in many downstream applications. However, the structure of these representations is often influenced by lexical overlap rather than semantic content. Our understanding of the relationship between this lexical influence and semantic content, and its implications for downstream tasks, remains limited. In this work, we investigate representations to quantify the effect of lexical overlap relative to semantic content. We consider several adversarial semantic stress tests and further connect our findings to the information theory perspective. We find that lexical influence extends across the depth of models, consistently across architectures, training regimes, and objective functions, including the models trained for semantic similarity. Moreover, we observe a mid-depth region in which both lexical and semantic signals degrade simultaneously, indicating a transitional regime where representations are poor for both surface form and meaning. We further demonstrate the effect of lexical influence on downstream uses of LLMs using summarization and model editing as a case study.

arXiv:2606.02724v1 Announce Type: new Abstract: Audio-visual speaker tracking aims to localize and track active speakers by leveraging auditory and visual cues, enabling fine-grained, human-centric scene understanding. This capability is essential for real-world applications such as intelligent video editing, surveillance, and human-computer interaction. However, existing datasets are largely limited to simple or homogeneous audio-visual scenes with coarse annotations. Such oversimplified settings bias evaluation toward static audio-visual co-occurrence, rather than rigorously assessing robust spatiotemporal modeling and cross-modal reasoning in complex, dynamic scenes. To address these limitations, we introduce AVTrack, a human-centric audio-visual instance segmentation (AVIS) dataset designed for dynamic real-world scenarios. AVTrack features diverse and challenging conditions, including camera motion, visual occlusions, and position changes. Evaluations of representative AVIS methods on AVTrack reveal substantial performance degradation, establishing AVTrack as a challenging benchmark for robust human-centric audio-visual scene understanding in complex environments. We further provide a simple yet effective baseline to facilitate future research. Project website: https://FudanCVL.github.io/AVTrack/

arXiv:2606.02776v1 Announce Type: new Abstract: When large language models (LLMs) are used in high-stakes scenarios, such as legal, medical and financial advice, even a single conversation history is enough to drive differences in outcomes between users. Prior work has demonstrated that this results in outcome disparities between sociodemographic groups, with some groups receiving more advantageous outcomes than others. In this work, we demonstrate that LLMs actually struggle to infer user sociodemographics from a single conversation history and that although there are disparities between sociodemographic groups, they are minimal in magnitude. To investigate what the main driver of these disparities is, we compare user sociodemographics to a range of (psycho)linguistic features of conversations, including conversation topic, emotions, and readability. We find that conversation topics are most predictive of LLM-generated advice within a conversational context, which, to some extent, function as proxies for sociodemographic groups and often affect advice in unpredictable ways. This is cause for concern and highlights the need for future research to better understand and, if needed, mitigate the effect of conversational context on LLM outputs in high-stakes scenarios.

arXiv:2606.02780v1 Announce Type: new Abstract: The success of the transformer architecture as the backbone of modern LLMs is in large part due to its use of attention layers. An attention layer follows the standard neural network paradigm: it takes the residual stream as input and thereby produces context-dependent query, key, and value vectors. However, we find that model performance meaningfully improves when deeper layers learn only a context-free value vector to preserve the original token information, without drawing on any context from the residual stream. When the model has access to this context-free value vector, adding back the context-dependent component provides little additional benefit for aggregate benchmark performance. Such context-free value vectors can be stored as sparse model parameters, eliminating the need to recompute or persistently cache these values. Through systematic ablations on the key design choices for such context-free value vectors, we propose Bank of Values (BoV), a new way of computing value vectors in attention by learning a lookup table of token-specific value vectors for each of the last third of layers. Across 135M and 780M models, BoV improves validation loss over standard attention and, at 780M, the average score across 21 benchmarks, matching the previous best method that adds token information to the value vector with less compute and memory.

arXiv:2606.02595v1 Announce Type: new Abstract: Dynamic pricing in short-term rental (STR) markets presents a distinctive challenge for online learning algorithms: pricing decisions carry significant financial risk, operators require explainability, and market feedback is sparse (one booking outcome per listed night). We introduce the Human-in-the-Loop Gated Bandit (HITL-GB) framework, in which a contextual bandit algorithm generates price recommendations but a human agent retains authority to accept, modify, or reject each recommendation before it is applied. We show that under this approval constraint, historical pricing data -- collected under a prior deterministic policy -- is structurally equivalent to on-policy warm-up data for initialising the bandit's posterior, bypassing the weeks-to-months cold-start period that renders pure online bandit learning impractical in sparse-feedback markets. We formalise the approval-gated reward signal, derive a regularised ridge-regression warm-up procedure from historical episodes, and validate the approach on real STR production data (anonymised urban market, 2 rooms, April 2022 -- April 2026, 1,461 nightly pricing episodes). Our warm-up procedure compresses effective cold-start from ~150 episodes to ~30 episodes when initialising agents from the Hierarchical Factored Thompson Sampling (HF-TS) family. We further argue that the structural equivalence result is domain-agnostic: any high-stakes domain where human approval is legally or operationally required -- including clinical drug dosing, credit origination, content moderation, and radiological diagnosis -- satisfies the same conditions and benefits from the same warm-up strategy. In regulated industries, mandatory human oversight is thus a statistical asset rather than a deployment constraint.

Daniele Levis Pelusi/Unsplash, CC BY-NC-SA Australia’s data centre rush now rivals the mining boom. OpenAI chief executive Sam Altman last week said Australia could become a “data centre capital of the world”. This would come at an environmental cost. Water use is a common concern. One report estimates AI centres could use billions of litres of water a year. But what do the numbers say? Based on the value derived per megalitre, data centres look less threatening and more likely to be a highly economically productive use of water. The bigger problems are energy and location. As a new report suggests, electricity demand from data centres could outstrip clean power from renewables and lead to new gas plants. Before committing fully, we need granular detail on how much water and energy these centres use. What value do we get from water? In the 2023–24 financial year, Australian industries consumed about 17.6 million megalitres of water – about 30 times the water in Sydney Harbour. Of this, agriculture, forestry and fishing consumed about two-thirds of the total – nearly 11.8 million ML. This water was used to produce goods valued at A$54.6 billion – roughly $4,600 for every megalitre consumed. Compare this to “other industries”, the category covering data centres. A megalitre of water in this sector was valued at $2.3 million – 500 times more value than if used on a farm. How much water do data centres use? We can only make a rough estimate on water use due to a lack of clear data. Research shows data centres need about 25ML of water per megawatt of capacity. Australia has about 300 data centres with about 1.3 gigawatts of operating capacity. Using these figures, Australia’s current data centres would use 15,000–35,000ML a year. That would be a fraction of 1% of the water used nationwide – close to a rounding error. There are three caveats. First, credible estimates of water use vary widely. Second, most estimates – including this article – only count water used directly for cooling. Data centres can be remarkably frugal with this water and getting more efficient. But data centres indirectly use substantially more in the water used to produce the electricity powering them. Coal, gas and hydro plants all need water. Third, proposed new data centres are much, much larger than existing ones. Some are seeking between 5ML and 40ML a day. Sydney is set for huge growth in AI data centres. If all 41 in the pipeline or under assessment are built, they would directly use 15–20% of Sydney’s water supply within a decade. Sydney would bear the strain on water supplies in return for an upfront economic benefit from construction and some ongoing jobs. But the economy wide $116 billion boost to GDP from AI adoption by all industries over the next ten years would be spread nationally. It would make sense to locate data centres where water is more abundant and cheaper. Energy is a bigger concern than water At present, data centres use just over 2% of the electricity on the National Electricity Market. This would almost triple to 6% within four years, according to Australian Energy Market Operator forecasts. The Clean Energy Finance Corporation estimates the figure could be 11% within a decade. Energy use isn’t inherently bad. What matters is whether increasing demand will be met by renewables – or gas. Data centres are hungry for energy. The question is how that demand will be met. Westend61/Getty We need better data – on data centres We can’t manage what we don’t measure. Data centres are a textbook example of a data gap impeding good policy. The Australian Bureau of Statistics rolls data centres into a broader category. This means we can’t access detailed statistics on how much water or energy data centres use. Nor how much they add to the national accounts. The federal government has introduced new expectations on water use and efficiency for data centre operators. That’s something. But it’s not the same as a national picture that fits with existing official statistics. Only one data centre meets the new national water-efficiency rating. Surprisingly, Australia’s National AI Plan has little focus on water and energy. State and federal water ministers have named data centres as an emerging threat to water security. A Senate inquiry is in progress. We need to track water use better Australia’s water accounts measure how much water every industry uses. But they don’t track how much water is lost to evaporation or value all water used. Water supply and sewerage are bundled together in even the most detailed view of the national accounts, meaning neither can be seen clearly. So while data centres appear to be a high-value use of water, we can’t confirm it. There are signs of change. Australia uses the international System of Environmental-Economic Accounting to track water use. This is being rewritten now, in part to address these issues. The national accounts have also begun treating damage to nature as a cost of production. Both shifts matter, no longer treating the environment as free lunch. To finish the job, authorities will have to properly value all the water used by industries and disentangle data centre data from other industries. This would turn a noisy debate into a measurable one. Time to keep tabs on AI Based on the data we do have, we can say Australia’s data centre boom is neither the water villain some fear nor the cost-free miracle its promoters describe. Instead, it looks like a high-value industry arriving at record speed which is relatively light on water use and fairly heavy on energy. With better data in hand, the numbers – not the headlines – should decide where the next megalitre and the next megawatt should go. Michael Vardon is a member of the Australian government's Technical Advisory Panel for Environmental-Economic Accounting and is on a United Nations team working on environmental-economic accounting.