LLMJacking is Costing Some Businesses Up to $1m, Warns Google
Google Mandiant has warned of the high cost of hijacked access to large language models, or LLMJacking
Google Mandiant has warned of the high cost of hijacked access to large language models, or LLMJacking
A new approach known as “reverse game theory” aims to control powerful AI models by designing systems that incentivize them to stick to the rules. As AI models become more capable, safety researchers are scrambling to ensure they are…
Cisco combines network intelligence with OpenAI models to strengthen enterprise security. Learn how it is building more adaptive defenses.
The company will use the capital to expand its AI team and develop its Large Logistics Spatio-Temporal Model, which it uses to improve delivery allocation and r
Models are commodities, harnesses are the valuable part
Accessing external memory puts the brakes on physical AI, so carmaker has found an economical and speedy shortcut
Some Goldman traders had used AI coding tools to model market moves and create graphic content.
With the accelerating intelligent transformation of global manufacturing, industrial robots have become critical equipment in modern smart factories. However, the growing prevalence of multi-variety, small-batch production has made the high programming threshold, tedious reconfiguration, and limited environmental adaptability of traditional industrial robots increasingly prominent. Over the past few years, large language models (LLMs) have proven to have strong semantic…
This paper presents a trajectory tracking control strategy for non-holonomic mobile robots with unavailable orientation angle measurements. An observer is developed to reconstruct the trigonometric terms required by the controller, while the Fourier-series-based reinforcement learning scheme performs online estimation of unknown nonlinear terms, and the controller compensates for bounded external disturbances. The nonlinear system is transformed into lower dimensional…
The behaviour of humans when collaborating with other actors such as other humans or robots has been studied for decades and remains critical for unlocking mixed human-robot team performance. To achieve better, more effective human-robot collaboration (HRC) in industrial settings, a robot needs to have robust models to predict human behaviour when working towards a shared goal. Inferring and modelling human intentions is expected to play a pivotal role in accurately…
Nuclear decommissioning increasingly relies on digital technologies to support characterisation, planning, monitoring, and decision-making in environments where access is limited by radiation, contamination, or other operational constraints. Digital twins provide a framework for integrating geometric, radiological, and operational information, while robotic systems acquire data and perform tasks with reduced human exposure. Although often treated separately, the two are…
Autonomous mobile robots (AMRs) that depend on remote inference introduce communication and service availability into the navigation control path. This paper presents a cloud-independent mapless-navigation architecture integrating visual features distilled offline from a vision–language model, PointNet++ light detection and ranging (LiDAR) features, and a Twin Delayed Deep Deterministic Policy Gradient (TD3) controller. An 8-bit integer (INT8)-quantized Vision Transformer…
To address the problems of cable slackness and non-smooth cable-tension variation curves during the motion of an antagonistic cable-driven parallel robot (CDPR), this paper focuses on a three-degree-of-freedom (DOF) redundantly constrained CDPR. First, an inverse kinematics model of the robot is established using the vector closed-loop principle. Second, a dynamics model of the robot is established using the Newton–Euler method. Linear programming, quadratic programming, and…
Purpose This study develops a fabrication-aware computational workflow for architectural applications using Wire-and-Arc Additive Manufacturing (WAAM). It addresses the gap in methods that integrate computational geometry with fabrication constraints, aiming to produce geometries that are both generative and manufacturable. The research formalizes a parameter-driven design-to-fabrication framework bridging computational modeling and robotic metal deposition, providing a…
To address the problems of inadequate anomaly representation in IoT traffic and poor anomaly localization capabilities, this study proposes a novel anomaly detection and localization framework based on probabilistic autoencoder reconstruction, temporal alignment, and multi-scale temporal feature learning. The model uses a Variational Autoencoder to learn the distribution of normal traffic and performs sequence reconstruction. It combines Dynamic Time Warping to improve the…
Precise mixed-crop recommendation is essential for increasing agricultural productivity, improving land utilization, and supporting sustainable farming. Effective crop combinations depend on several environmental and soil factors, making accurate recommendations a challenging task. This study compared classical and hybrid machine learning approaches for mixed-crop recommendation using an indigenous dataset from Andhra Pradesh, India. The dataset consisted of seven key…
AI-enabled vision-based perception models are increasingly used in precision agriculture for tasks such as fruit detection, trunk detection, and autonomous navigation. As these models move into field deployment, their errors no longer stay confined to the image: they propagate into the counts, maps, and yield estimates that downstream automation depends on, making adversarial robustness a security concern. Yet this robustness remains underexplored in orchard settings. Using…
Computational drug discovery is increasingly transforming the way therapeutic candidates are identified, prioritized, and optimized, with advances in artificial intelligence (AI), machine learning, molecular modelling, cheminformatics, and data-driven approaches expanding the possibilities for early-stage drug discovery. The growing availability of open-source software, public chemical and biological databases, cloud and high-performance computing, and increasingly…
Introduction Metabolic reprogramming and immune evasion are tightly linked in gastric cancer (GC), yet metabolic signatures are rarely leveraged to guide treatment decisions or herbal small molecule drug discovery. Methods We profiled glycolysis, lipid, and tyrosine metabolism pathways across the TCGA-STAD and GSE62254 cohorts, identified prognostic metabolic subtypes, and distilled these findings into a three-gene risk signature ( VCAN, NOS1, LCT ). The signature was…
Artificial intelligence (AI) systems increasingly shape organisational decision-making, yet, the institutional capacities required to align these technologies with Brazil’s General Data Protection Law (LGPD) remain uneven. This article argues that many organisations, whether public or private, face institutional consciousness gaps, defined here as structural, cognitive, and cultural deficiencies that hinder their ability to understand and operationalise the ethical and legal…
ABSTRACT Why do first‐mover countries that once championed open trade now impose strict export controls on advanced technologies? And how can developing economies build R&D capacity when cut off from frontier innovations? This paper addresses these questions by developing a general equilibrium model featuring endogenous specialization in intermediate goods and the international division of labour. Allowing institutions and policies to adjust contingently in pursuit of…
Abstract The globalization of the semiconductor supply chain has intensified the proliferation of counterfeit integrated circuits (ICs), posing severe threats to the reliability and security of safety critical systems. Federated Learning (FL) has emerged as a promising privacy preserving paradigm for collaborative counterfeit IC detection, enabling distributed entities to jointly train machine learning models without exchanging sensitive raw data. However, the decentralized…
This study proposes a machine learning-enhanced ARIMA-LSTM framework for predicting monthly housing prices across the twenty-two districts of Tehran, Iran. The model combines ARIMA-based trend stabilisation with LSTM-based nonlinear learning and uses a data-driven machine-learning integration stage to improve city-wide price prediction. To extend the city-wide forecasts to local markets, a district ratio modelling mechanism is used to reconstruct district-level prices from…
Introduction Since the 1990s, East Asia has prioritized the production of scientific knowledge based on monitoring and modeling data, devoting significant effort to understanding the source-receptor relationships of transboundary air pollution among countries in the region. However, the lack of a common understanding of the causal mechanisms underlying air pollution in the region has persisted for decades. Despite persistent disagreements regarding the causal understanding…
Automating compliance analysis against catalogs such as NIST SP 800-53, which defines over a thousand hierarchical security and privacy controls, is attractive for resource-constrained organizations, yet sending sensitive infrastructure documents to cloud large language models is an unacceptable confidentiality risk. The natural alternative – local retrieval-augmented generation (RAG) on commodity hardware – raises a deployment question: given a privacy-mandated local RAG…
This study examines hypothetical cybersecurity-application utilisation interventions for Computer Science undergraduates in Enugu State, Nigeria. Five author-supplied datasets, each containing 364 cases, provide audited sex-specific calibration moments for a hybrid agent-based and system-dynamics model. Seven portfolios are compared over 24 months using 400 paired replications; additional experiments examine dose, effect assumptions, targeting, academic calendars, turnover…
Document classification in the cybersecurity domain is essential for the efficient identification of emerging threats, technologies, and policy trends. However, existing datasets and language models are biased toward technical, attack-oriented texts such as CTI reports and CVE descriptions, and there is a critical lack of data for the cybersecurity policy domain. To address this gap, we introduce the first benchmark dataset for cybersecurity policy document classification…