AI

Anthropic Keeps Claude Mythos AI Private After It Discovered Thousands of Critical Security Flaws

Anthropic limits Claude Mythos rollout as it spots vulnerabilities in many software systems

Japan relaxes privacy laws to make itself the ‘easiest country to develop AI’

Anthropic limits access to AI model, fearing future of cyberattacks

Your LLM Agent Can Leak Your Data: Data Exfiltration via Backdoored Tool Use

The Planetary Cost of AI Acceleration, Part II: The 10th Planetary Boundary and the 6.5-Year Countdown

Synthetic Trust Attacks: Modeling How Generative AI Manipulates Human Decisions in Social Engineering Fraud

Anthropic Teams Up With Its Rivals to Keep AI From Hacking Everything

A new Anthropic model found security problems ‘in every major operating system and web browser’

Intel will help build Elon Musk’s Terafab AI chip factory

CertiK unveils AI Auditor to improve early detection of blockchain vulnerabilities

Anthropic secures multi gigawatt TPU deal with Google and Broadcom

Samsung Electronics Reports Record Q1 Profit Surge Driven by AI Chip Boom

V-Reflection: Transforming MLLMs from Passive Observers to Active Interrogators

Researchers introduce V-Reflection, a new framework that transforms Multimodal Large Language Models (MLLMs) from passive observers to active interrogators through a 'think-then-look' mechanism. The approach addresses perception-related hallucinations in fine-grained tasks by allowing models to dynamically re-examine visual details during reasoning, showing significant improvements across six perception-intensive benchmarks.

Preserving Forgery Artifacts: AI-Generated Video Detection at Native Scale

Researchers developed a new AI-generated video detection framework using a large-scale dataset of 140K videos from 15 generators and the Qwen2.5-VL Vision Transformer. The method operates at native resolution to preserve high-frequency forgery artifacts typically lost in preprocessing, achieving superior performance in detecting synthetic media.

Artificial Intelligence and Systemic Risk: A Unified Model of Performative Prediction, Algorithmic Herding, and Cognitive Dependency in Financial Markets

A new unified model demonstrates that AI adoption in financial markets creates systemic risk through three channels: performative prediction, algorithmic herding, and cognitive dependency. Using SEC Form 13F data from 2013-2024, researchers found AI adoption generates superlinear growth in systemic risk and tail-loss amplification of 18-54%.

SoLA: Leveraging Soft Activation Sparsity and Low-Rank Decomposition for Large Language Model Compression

Researchers propose SoLA, a training-free compression method for large language models that combines soft activation sparsity and low-rank decomposition. The method achieves significant compression while improving performance, demonstrating 30% compression on LLaMA-2-70B with reduced perplexity from 6.95 to 4.44 and 10% better downstream task accuracy.

Customized User Plane Processing via Code Generating AI Agents for Next Generation Mobile Networks

Researchers propose using generative AI agents to create customized user plane processing blocks for 6G mobile networks based on text-based service requests. The study evaluates factors affecting AI code generation accuracy for network-specific tasks, finding that AI agents can successfully generate desired processing functions under suitable conditions.

Robust LLM Performance Certification via Constrained Maximum Likelihood Estimation

Researchers propose a new constrained maximum likelihood estimation (MLE) method to accurately estimate failure rates of large language models by combining human-labeled data, automated judge annotations, and domain-specific constraints. The approach outperforms existing methods like Prediction-Powered Inference across various experimental conditions, providing a more reliable framework for LLM safety certification.

Evolutionary Search for Automated Design of Uncertainty Quantification Methods

Researchers developed an LLM-powered evolutionary search method to automatically design uncertainty quantification systems for large language models, achieving up to 6.7% improvement in performance over manual designs. The study found that different AI models employ distinct evolutionary strategies, with some favoring complex linear estimators while others prefer simpler positional weighting approaches.

Is your AI Model Accurate Enough? The Difficult Choices Behind Rigorous AI Development and the EU AI Act

A research paper challenges the common view of AI accuracy as purely technical, arguing it involves context-dependent normative decisions that determine error priorities and risk distribution. The study analyzes the EU AI Act's "appropriate accuracy" requirements and identifies four critical choices in performance evaluation that embed assumptions about acceptable trade-offs.

Large Language Models Align with the Human Brain during Creative Thinking

Researchers found that large language models align with human brain activity during creative thinking tasks, with alignment increasing based on model size and idea originality. Different post-training approaches selectively reshape how LLMs align with creative versus analytical neural patterns in humans.

The Persuasion Paradox: When LLM Explanations Fail to Improve Human-AI Team Performance

Research reveals a 'Persuasion Paradox' where LLM explanations increase user confidence but don't reliably improve human-AI team performance, and can actually undermine task accuracy. The study found that explanation effectiveness varies significantly by task type, with visual reasoning tasks seeing decreased error recovery while logical reasoning tasks benefited from explanations.

Incompleteness of AI Safety Verification via Kolmogorov Complexity

Researchers prove a fundamental theoretical limit in AI safety verification using Kolmogorov complexity theory. They demonstrate that no finite formal verifier can certify all policy-compliant AI instances of arbitrarily high complexity, revealing intrinsic information-theoretic barriers beyond computational constraints.

PassiveQA: A Three-Action Framework for Epistemically Calibrated Question Answering via Supervised Finetuning

Researchers propose PassiveQA, a new AI framework that teaches language models to recognize when they don't have enough information to answer questions, choosing to ask for clarification or abstain rather than hallucinate responses. The three-action system (Answer, Ask, Abstain) uses supervised fine-tuning to align model behavior with information sufficiency, showing significant improvements in reducing hallucinations.