The Independent reveals a Facebook page called 'Life in Britain' posting AI-generated divisive videos about UK politics, managed from Sri Lanka. Despite being flagged a month ago, Facebook has taken no action. Experts warn of a foreign-influenced AI campaign eroding trust.

The CAI Dataset is a fourteen-month corpus of cybersecurity LLM trajectories collected via the open-source CAI agent framework. It includes 230,935 session logs and 26,027,742 user prompts from 16,768 source IPs across 123 countries, addressing the finding that expert operator trajectories, not base-model capability, are the bottleneck for cybersecurity LLM performance. The dataset reveals that operators routinely paste live credentials and tokens into prompts, concentrating sensitive context within a few cloud API providers, posing a systemic risk. The paper advocates for on-premise, privately-hosted cybersecurity-specialized LLMs to balance productivity and confidentiality.

A technical review before launch is critical for AI-built apps. Check code ownership, prepare for the 80% build limit, secure user data, and get a second opinion. Builder.ai's bankruptcy highlights the gap between a demo and a production-ready product.

OpenClaw has grown from a weekend project into a global movement, with 4.5 million new claws born every week and the fastest-growing repository in GitHub history. Today, it announces the formation of a non-profit foundation to steward the project as open and independent. The foundation will provide governance, stable funding, and a full-time team. Partnerships with OpenAI, NVIDIA, Microsoft, and the University of Michigan aim to advance personal AI agents.

SpaceXAI (xAI) publicly launched Grok 4.5, a coding- and agent-focused frontier model positioned as Opus-class but faster, more token-efficient, and lower cost. Trained in partnership with Cursor, it is priced at $2/M input and $6/M output tokens with a 500k context window (expanding to 1M soon). Independent evaluations highlight its efficiency, ranking #4 on the Artificial Analysis Intelligence Index and offering strong cost-performance tradeoffs.

The quality of AI code generation depends on the consistency of the codebase. Popular stacks benefit from AI's training data, while inconsistent or proprietary codebases require more context, raising costs and lowering quality. Rewriting software to adopt clear patterns can flip this dynamic, offering a competitive edge.

An undercover investigation by Correctiv reveals that Germany's far-right AfD party has developed Alternita, an AI software suite using Google Gemini, OpenAI, and Anthropic Claude to generate provocative social media posts known as 'rage bait', aiming to control messaging and maintain online dominance.

Compendium is a shared workspace for teams and AI agents, developed by Cerenovus, which brands itself as 'The AI Company Brain'. The tool facilitates human-AI collaboration to boost productivity.

A cautionary tale about shadow governance, AI, and the quiet weaponization of ambiguity. A security expert joins an AI startup, strictly follows an unpublished policy, but gets his reputation ruined by rumors for using AI tools, while real violators go underground. The company eventually releases a permissive policy but with comprehensive monitoring.

This article outlines six steps to identify and remove unnecessary tool definitions, instructions, and other bloat from Claude Code's payload, reducing token usage and cost. Using /context, a logging proxy, and configuration flags/deny rules, users can significantly trim the overhead.

This article explains how to reproduce an indirect prompt injection attack on a Retrieval-Augmented Generation (RAG) pipeline, detailing the attack principle, implementation steps, and defense recommendations.

GemNav is a novel visual robot navigation policy that adapts a frozen Multimodal Large Language Model (MLLM) via Low-Rank Adaptation (LoRA) on the language tower alone, without auxiliary visual encoders or continuous regression heads. It uses a shared discrete token vocabulary for waypoints and navigation signals, and a soft-decoded auxiliary loss recovers metric structure. Trained on just 8.7 hours of data, it zero-shot transfers to four unseen environments, stopping within 0.25-0.42m of goals across 20 trials. Results indicate discrete-token adaptation of frozen MLLMs is a data-efficient, deployable alternative for robot navigation.

CaLiSym extends exact symplectic learning to non-conservative robotic systems by imposing geometric priors on a structured lifted canonical phase space. It uses an explicit algebraic lift, avoiding recurrent or ODE integration, and introduces GRB-SympNet. Experiments show improved out-of-distribution prediction on a dissipative double pendulum, real-world quadrotor, and contact-rich quadruped while preserving symplectic form.

This paper proposes a continual learning-based control framework for modular soft robots that incrementally adapts to morphology changes without forgetting prior knowledge, validated in simulation and on a real robot.

EvoPlan is a neuro-symbolic framework combining LLM fluency and classical PDDL planner guarantees. It has three parts: offline mining of global STL constraints from demonstrations, an evolutionary PDDL planner, and a constrained execution loop. All LLM calls use local open-weight models, enabling on-robot deployment without cloud dependency. Tested on Bench2Drive, HA-VLN-CE, and ALFWorld benchmarks.

This review synthesizes 183 contributions from 2017-2026 covering VLA architectures, training recipes, action representations, bimanual coordination (2022-2026), UAV navigation and control (2017-2026), language grounding, and cross-cutting concerns. It shows that strategies from bimanual VLAs transfer to aerial systems and identifies fourteen research directions.

This paper proposes CILC, a system that uses Secure Multi-Party Computation (SMPC) to detect loop closure candidates in multi-agent SLAM without exchanging global descriptors in the clear, protecting against compromised agents. Experiments validate real-time performance on visual and LiDAR descriptors with reduced information leakage.

RoboSnap is a real-to-sim framework that turns a single RGB image into a simulation-ready scene using a layered design: collision-aware foreground assets for stable robot interaction and 3D Gaussian splatting for faithful background appearance. Experiments on DROID scenes and real-robot tasks show reliable trajectory replay, task-specific synthetic data generation, and meaningful sim-real correlation. The work also introduces DROID-Sim, a companion dataset of 564 real-world scenes.

NativeMEM is a novel VLA model that compresses historical frames into single tokens using native memory compression, eliminating external memory modules. It boosts success rates from 32.4% to 84.0% in simulation and up to 98.7% on real robots, using only 20% training data.

Pelican-VLA 0.5 is a unified Vision-Language-Action (VLA) model that integrates vision-language understanding, future-frame generation, and action prediction within a single architecture. It achieves attention-level generalization through learnable Reasoning Slots inserted between perception and action, enabling the action pathway to focus on instruction-relevant objects and contact regions without explicit supervision. This behavior persists across unseen scenes and robot embodiments, outperforming other open-source VLA baselines.

This paper proposes ProMoE-FL, a prototype-conditioned mixture-of-experts framework for robust missing-modality feature synthesis in multimodal federated learning. It builds a global client-aware prototype bank capturing clinically meaningful modality priors across institutions, and uses direction-aware expert routing to dynamically synthesize missing features. Experiments on four chest X-ray datasets show consistent outperformance over state-of-the-art methods in both homogeneous and heterogeneous settings.

Video Diffusion Models (VDMs) achieve high quality but are computationally expensive. Recent few-step distillation accelerates inference but ignores varying computational demands across noise levels. This paper proposes a post-training framework that integrates dynamic structural sparsification into distillation, jointly optimizing denoising steps and model sparsity to create a step-specific Mixture-of-Models (MoM). A Progressive Training Strategy with Output Rollout Mechanism ensures stability, and a specialized inference engine enables efficient deployment. On Wan-14B, it removes 24% per-step FLOPs on top of 4-step distillation, achieving 1.2x wall-clock gain and 30x speedup over the 50-step teacher with competitive quality.

SpaR3D-MoE is an end-to-end framework enabling adaptive 3D spatial reasoning from sparse RGB inputs via adaptive spatiotemporal manifold sampling and a geometry-inductive mixture-of-experts. It achieves SOTA on VSI-Bench (63.5 average, +7.8 over baseline), with 35.4% and 51.4% improvements on Route Plan and Relative Direction tasks respectively.

This paper introduces the Difficulty-Aware Medical Instructional Video Question Answering (DA-MIVQA) shared task for NLPCC 2026. It extends previous benchmarks by explicitly distinguishing questions based on the type and complexity of evidence required. Simple questions can be answered from subtitle text, while complex questions require visual grounding, procedural understanding, and cross-modal integration. Three tracks are included: DA-TAGSV, DA-VCR, and DA-TAGVC. The dataset is collected from public medical instructional channels, covering first aid, emergency response, rehabilitation, nursing, and general medical education, with manual difficulty annotations.

This study investigates the relationship between counterfactual fairness (CF) and group fairness (GF) in image classification. By constructing new datasets with high-quality image editing, it finds that CF does not imply GF in images, contrary to tabular data results. The discrepancy is attributed to a latent attribute correlated with the sensitive attribute. The proposed Counterfactual Knowledge Distillation (CKD) method reduces reliance on this attribute, allowing CF-achieving models to also satisfy GF.

Line segment detection is key in visual SLAM, 3D reconstruction, and industrial inspection. Deep learning models require several megabytes of memory, exceeding MCU capacity. This work proposes MiLSD, a detector for sub-megabyte budgets. It compares three output representations in a compact backbone, finding F-Clip (center with length and angle) most effective. 8-bit quantization preserves full precision; 4-bit degrades angle regression. With 1 MB activation budget, MiLSD improves sAP10 from 10.6 to 24.1 on ShanghaiTech Wireframe. The paper maps accuracy-memory trade-offs for embedded vision.

This paper introduces LipSSD, a Lipschitz-constrained Single Shot MultiBox Detector (SSD) that improves adversarial robustness by design. On Pascal VOC, adversarially trained LipSSD improves mAP@50 on unseen attacks by up to 15 points over standard adversarially trained SSD, and shows robustness gains on safety-critical datasets LARD and KITTI while largely preserving clean performance.

This study fine-tunes two pretrained latent diffusion models, Protogen v3.4 and Stable Diffusion v1.4, on a curated dataset of high-resolution Ulos motifs to generate culturally consistent yet novel designs. Protogen v3.4 significantly outperforms Stable Diffusion v1.4 in terms of FID and IS, highlighting a fidelity-diversity tradeoff. A guidance scale of 5–9 is recommended for optimal balance.

CoFINN is a physics-informed deep learning framework for predicting compressible flow fields, embedding finite-volume conservation physics directly into the training process. It outperforms traditional data-driven CNNs and classical physics-informed methods, reducing drag prediction error by up to 34% on transonic airfoil flows and 15% on average, especially beneficial in limited-data regimes.

The paper proposes a unified deep learning pipeline integrating semantic segmentation, regression-based severity estimation, and disease classification for plant disease severity quantification. On the Apple Tree Leaf Disease Segmentation dataset, U-Net with MobileNetV2 achieves 98.20% pixel accuracy, 0.70 mIoU, and 99.41% detection accuracy at 14.7 ms per image, suitable for real-time use. The computed severity index strongly correlates with expert annotations (r=0.968), demonstrating reliability for automated crop monitoring.