AI News Brief Daily Summary 2026-09-17

200 posts published today

21:33Taming the Agentic RAN: Stability-Guaranteed Arbitration of Autonomous AI Agents in O-RAN
21:33Ask the Tool, Don’t Guess: Agent Tool Calls Hold Their Progress, and the Serving System Should Read It
21:33Decodable but Misrouted: Sparse Features Uncover a Readout Gap in Vision-Language Models for Harmful Meme Detection
21:33Microsoft Open-Sources TauGrid: A Kubernetes-Native Stack for GPU AI Workloads
21:33ASLEval: Measuring Privacy Exposure Displacement in LLM Agent Sessions
21:32OpenAI Introduces Astra for Law With Legal Search and Trusted Access
21:32GrainSpeech: Less Context, More Detail for Compact Speech Synthesis
21:04Rethinking Critic Learning in PPO: Understanding and Mitigating Value Flattening
21:04UN System Data Commons Launches as AI-Ready Global Statistics Platform
21:04The fix for rogue AI agents could be more AI
21:03Reimagining advertising with AI
21:03Using OCR Heads to Verbalize Image Semantics
21:03Anthropic Launches Claude Code Projects in Beta: Parallel Cloud Sessions That Keep Running After You Close Your Laptop
21:03Echo: Learning-based Matching Decompilation using Trusted Back Translation
21:03Google Labs Expands CC Into an AI Agent for Families and Households
21:03A Scalable Framework for Automated NER Annotation Correction in Low-Resource Languages
21:03OpenAI caught its models leaving notes to successors to hide bad behavior
21:03ProgramDistill: From Interactive Web Apps to Verifiable Reference-Guided SWE Tasks
21:00AI News Brief Hourly Summary 2026-09-17 23h : 21 posts
20:33Beyond EER: Multi-Dimensional Evaluation of Information Leakage in Speaker De-Identification
20:33CoRe-MARL: Cooperative Redistribution Under Unknown Dynamics Using Recurrent Multi-Agent Reinforcement Learning
20:33UN turns to Google to make its global data ready for AI agents
20:32GenStream: Semantic Streaming Framework for Generative Reconstruction of Human-centric Media
20:32Introducing Astra for Law
20:32Generalist-Specialist Mixture-of-Experts for Rare Pathology Detection in Multimodal Imaging
20:32Is the AI safety debate about safety or control?
20:32PACT: Can Enterprise AI Assistants Be Trusted Under Pressure?
20:04How to Build Effective Evals for AI Agents
20:04Making global data easier to explore
20:04Anthropic Reports Claude Optimized 30+ Open-Source Biomolecular Models
20:04On-the-Fly Homographies Calibration for Multi-Camera Tracking
20:03Microsoft exec called AI scraping ‘the largest theft of labor in human history,’ new unredacted filings reveal
20:03Interpretable Patch-Based Deep Learning for Wildfire Spread Prediction from Ensemble Simulations
20:03Building the materials foundation for AI
20:03Label-free steering: Compressing test-time reinforcement learning into bias-only subspaces
20:03Daniel Liechtenstein, CEO and Co-Founder of Hypercore – Interview Series
20:03Online Robust Reinforcement Learning Through Monte-Carlo Planning
20:03GSA Extends Anthropic’s Claude OneGov Offer for Federal Agencies
20:03Hypothesis-Driven Autonomous Materials Synthesis with Multimodal LLM Agents
20:00AI News Brief Hourly Summary 2026-09-17 22h : 17 posts
19:32Multitask Reinforcement Learning for Assisting Choice Model Specification
19:32ActionPiece: Rethinking Action Tokenization for Autoregressive Vision-Language-Action Models
19:32MiST: Mid-Training LLMs for Cybersecurity
19:32OpenAI reportedly closes in on solving the Hodge conjecture, its second Millennium Prize Problem
19:32CSWAM: Better Causal Semantic Representations for Out-of-Distribution Generalization in World Action Models
19:32Anthropic Redesigns Claude Code Projects to Coordinate Agent Threads
19:32VoiceTrace: A Benchmark and Retrieval Framework for Who-Said-What Speech Retrieval
19:04A Non-Linear Neuron Based Detection of Isolated Pixels in Binary and Grayscale Images using Contrast Sensitive Receptive Fields
19:04Even the king of England has his hesitations about AI
19:04GYROval: A Robust Benchmark for Cultural Value Orientation in Large Language Models
19:04Anthropic keeps pushing Claude Code toward autonomous coding with new parallel agent workflows
19:04Semantic CSI Feedback for Beam Selection: When Task-Aware Embeddings from Sparse Pilots Outperform Full-Bandwidth Reconstruction
19:03Figure Introduces Helix 2.5, Tested Zero-Shot in 30 Unseen Homes
19:03TERN: A Delta-rule Memory with a Seasonal Reference and Online Adaptation for Epidemic Forecasting
19:03Lasso Study Finds Text Watermarking Shifts LLM Refusals and Tool Calls
19:03Reliable Virtual Sensing: A Multi-Domain Benchmark for Robustness Under Sensor Failures
19:00AI News Brief Hourly Summary 2026-09-17 21h : 22 posts
18:33Trajectory Learnability for Offline On-Policy Distillation with Imperfect Teachers
18:33Magentic Raises $18M Series A to Build an AI Workforce for Industrial Operations
18:33Knowledge-Graph Based Augmentation versus Retrieval Augmented Generation for Cultural-Related Question Answering
18:33Kris Beevers, CEO and Co-Founder, Netbox Labs – Interview Series
18:33A Study of the Reliability of Agentic AI-Generated Programs
18:33Delta Unveils Energy-to-Compute Infrastructure for NVIDIA DSX AI Factories
18:33Look Less, Hear Better: Jointly Rewarded GRPO for Streaming ASR
18:32AWS Launches Amazon Connect Talent for AI-Led Hiring at Scale
18:32Autonomy in Check: Governor-Mediated Adaptive Security at the Edge
18:04Black Kite’s 2026 Manufacturing & Distribution Ransomware Report: Manufacturing Remains Ransomware’s Top Target
18:04What’s Actually Inside 24,723 Tokens of a Search Result? We Broke It Down, Field by Field
18:04Introducing the Life Sciences Verification Program
18:04${M}^2$Tok: Multi-head Multi-codebook Discrete Action Tokenization for Vision-Language-Action Models
18:04Reduce time-to-hire for quality candidates with AI-powered Amazon Connect Talent
18:04Remembering Solomon Marcus
18:04Anthropic Launches Life Sciences Verification Program in Beta
18:04Quanta: A Self-Contained Python Library for Hybrid Retrieval over Quantised Embeddings, Lexical Indexes, and Knowledge Graphs
18:03What’s So Good About ChatGPT Work? Here’s What I Found
18:03I code or AI code: A comparative evaluation of AI-rated scores in classroom observations
18:03Crusoe Secures $3.9B Series F at $30.9B Valuation to Build AI Factories
18:03APGEM: Adaptive Policy-Guided Error Mitigation for Quantum Reinforcement Learning on a Real-World CVRP Case Study
18:00AI News Brief Hourly Summary 2026-09-17 20h : 16 posts
17:33MoRE: Mixture of Reused Experts
17:33Pinterest teases a new ‘Restyle’ feature that lets you redesign your room with AI
17:33CPR: Combining global composing, local performing and full-sequence refining in piano rendering with continuous autoregressive modelling
17:33AI Agents Will Make Business Context an Operating Asset
17:32Beyond Accuracy: How Procedural Traces Shift the Decision Criterion of LLM Overseers
17:32GPT-6 Astra crushes Pokemon, Factorio, and Fallout 3 then spirals into Minecraft potato farming after one bad Creeper
17:32A Lightweight CNN Integrated Compact Convolutional Transformer for Multi-Scale Feature Learning and reducing computational complexity for breast cancer mammography image detection and classification
17:32Base Labs launches an open-weight AI safety partnership with Hugging Face and Goodfire
17:32CapMap-MS-TTA: 3rd Place Solution for the MUMU Track of the 8th LSVOS Challenge at ECCV 2026
17:03PentestChain: A Cost-Aware, MCP-Orchestrated Framework for Automated Penetration Testing with Free-Tier LLMs
17:03Linguistic Triggers of Gender and Racial Bias in Open-Weight LLMs Applied to Recruitment
17:03A Comprehensive Review of Generative Physical Artificial Intelligence
17:03Rethinking How We Evaluate Methodological Progress in Health AI
17:03Amazon launches Alexa+ in India with Hindi support
17:03DualSQL: Text-to-SQL with Multi-Agent Reinforcement Learning
17:00AI News Brief Hourly Summary 2026-09-17 19h : 18 posts
16:33From a River in Gilead to the Inference Distributions of Large Language Models: Covert Dialect Bias and Linguistic Profiling at Scale
16:33Agora: Git as Shared Memory for Collective AutoResearch
16:33An Empirical Evaluation of Cost-Efficient Large Language Models on Algorithmic Programming Tasks
16:33Mask 2D-3D: Adaptive Dual-Masked Autoencoder Network for Image-to-Point Cloud Registration
16:33ChatGPT pioneer launches Jev model for programmatic logic
16:33Physics-Informed Neural Networks for Fast Multilayer Spectral Inversion of H{\alpha} 6562.8 A and Ca II 8542.1 A Spectra
16:04Walking the Score Manifold: Continuous-time Generative Dynamics on Learned Data Manifolds
16:04A serverless, data-driven Git metrics dashboard using Amazon Quick Sight
16:04Insilico Medicine Releases Open Longevity AI Toolkit in Cell Study
16:04Newer Is Not Fairer: Gender Stereotyping in Text-to-Image AI Across Model Generations
16:04Selecting a vector store for Amazon Bedrock Knowledge Bases
16:04Pinecone Open-Sources VQ-Bench Vector Quantization Framework
16:04Whom Do AI Agents Work For? Role Assignment Induces Sponsorship Bias in LLM Recommenders
16:03How MRH Trowe enabled secure self-service AI agents in financial services
16:03EDCT-Bench: Uncovering Faithfulness Gaps in VLMs via Explanation-Driven Counterfactual Testing
16:03A shared agentic platform for Wood Mackenzie, on Amazon Bedrock AgentCore
16:03The Attention Within: Consensus Dynamics in Selective State Space Models
16:00AI News Brief Hourly Summary 2026-09-17 18h : 19 posts
15:34Does AI Assistance Leave a Temporal Fingerprint? Detecting Overreliance in AI-Assisted Writing and Programming
15:34Implementing defense-in-depth authorization for MCP tools on Amazon Quick
15:34Sierra Earns AIUC-1 Certification After Independent Audit
15:33Who Judges Matters: Measuring Family-Conditioned Preference in LLM-as-Judge Panels
15:33Enhancing industrial safety AI with synthetic data on Amazon SageMaker AI
15:33PrimeScientist: Strategic Allocation of Research Effort in Autonomous Research
15:33How Patient Data Is Driving the Future of Healthcare
15:33AfriSyCo: Measuring Assertive Framing, Verification, and Wording Sensitivity Around African-Language Content
15:33AI Companies Argue There’s a Hidden Cost of Banning AI and Screens at Schools
15:33RoboVAD: A Large Cross-Domain Evaluation Benchmark for Anomaly Detection in Robotic Arm Manipulation Videos
15:04Adaptive hybrid coupling with operator inference, the overlapping Schwarz alternating method and reinforcement learning
15:04Learning Nuclear Structure with AI: Radii and Collectivity
15:04Single-Phase Direct Liquid Cooling Is Proven for the Next Decade of Ultra-Dense Compute
15:04Lexara-RF: Reference-Free Metrics for Evaluating Conversational Visual Analytics Agents
15:04Lightmatter Joins Open CPX MSA With Bidirectional Passage L20 Module
15:04Learning Multi-Humanoid Pickup and Transport via Decentralized Object-Centric Control
15:04GPT-6 Astra: Pokemon champion in 18 hours, potato farmer after one Creeper mishap
15:04Procedural Pretraining for Molecular Property Prediction
15:00AI News Brief Hourly Summary 2026-09-17 17h : 18 posts
14:35Principled Koopman Representations with Kalman Inference for Efficient Time-Series Prediction
14:35Google, Nvidia, and Anthropic want Emerald AI to find space on the grid for more data centers
14:35QiT: Quantum-Inspired Transformer for Visual Recognition Task
14:34Prior Labs Releases TabPFN-3.5: A Tabular Foundation Model That Beats the Winning Otto Kaggle Solution With Default Settings
14:34SAiFE-gym: Model-based Environments for Automated Market Making with Concentrated Liquidity
14:342 days left to exhibit at TechCrunch Disrupt 2026
14:34The Free Inference Dimension: Complexity Measure for Zero-Collision Navigation under Hypothesis Mixtures
14:34Huawei plans Q1 2027 launch of new AI chip as it takes on Nvidia
14:34Reflections on Trusting Trust, Revisited: Contaminating Self-Modifying AI Coding Agents with Poisoned Benchmarks
14:05Is Luke the Author of a Gospel and the Acts of the Apostles?
14:05Google, Nvidia and Anthropic want Emerald AI to find space on the grid for more data centers
14:05AI and Human Approaches to Mathematical Problem Solving
14:05An OpenAI model kept slipping prompt injections into its own notes, and researchers still aren’t sure why
14:05Information Set Emulation: Causal Certificates for AI Derived EHR Features
14:05OceanStor M900 Brings PB-Scale Context Memory to Huawei SuperPoDs
14:05When AI Generates Covariates: Causal Typing and Estimand Drift in Sequential Experiments
14:05Rival AI agents, Instinct and Meta’s Muse, both add the ability to make calls
14:04HINT-Plan: Human Intention-Aware Robot Task Planning in Context-Rich Environments using Vision Language Models
14:00AI News Brief Hourly Summary 2026-09-17 16h : 13 posts
13:34Evolution of US Oral Political Language
13:34Confidence Comes from Experience: Experiential Confidence Estimation from Reasoning to Agents
13:34One Size Does Not Fit All! Dynamic Retriever and Generator Selection for RAG
13:34CALOS: Control-Affine Lyapunov On-manifold Safety Layer for Safe Deep Reinforcement Learning for Quadrotors
13:34Industrializing Humanization
13:34REVERSAL-BENCH: A Reversibility Axis and Reset Oracle for Measuring the Reset-Free RL Cliff
13:04Accelerating Diffusion Sampling via Speculative Draft Trees
13:04Reflect, Revise, Reuse: Training-Free Skill Evolution for GUI Agents
13:04Rethinking Domain Specialization for Open-Ended Scientific Reasoning in Astronomy Language Models
13:04Private AI and Cloud Repatriation Offering Reaches More Than 30 Countries
13:04Scaling Articulated Rationales for MLLM-based Recommendation
13:04Nums AI Releases Causilo: A Tabular Foundation Model That Tops TabArena Among Single Models
13:04The Missing “I Don’t Know”: Why Three Reasoning-Reliability Findings Converge on Calibrated Abstention
13:00AI News Brief Hourly Summary 2026-09-17 15h : 15 posts
12:33Evolutionary Ensemble Search: Council-Guided Program Evolution with Persistent Memory
12:33Structure is not mechanism: high-gain gated-FFN rows across text and genomic foundation models
12:33Lecture notes on Physics Informed Neural Networks, Neural Operators, and their applications
12:33Decentralized Optimal Equilibrium Learning Over Dynamic Networks
12:32MIND Raises $72M Series B to Scale AI-Native Data Loss Prevention
12:32Where Grokking Happens: Distributed Utility and Fourier Recoding Without a Module Switch
12:04BLADE: ReliaBle Dynamic Hardware-Aware SNN-ANN Boundary SeLection for Event-BAseD Object DEtection
12:04Independence-System Realisations in Single-Source Unsplittable Flow
12:04Healthcare AI Metrics Are Missing the Point. Closing the Access Gap Is AI’s Greatest Return
12:04WARD: Runtime Workload-Adaptive Vision TRansformer Framework for Dependable Edge AI
12:045 Free Zoomcamps From Data Pipelines to AI Agents
12:04Pay Only for Disagreement: Certified No-Regression Verdicts for Model Updates with Matching Label-Complexity Bounds
12:04CleanSpark Proposes $2.227B Senior Secured Notes to Fund Sandersville Build-Out
12:04REQAP: Resilient Weight Packing and Quantization for Edge DNN Acceleration
12:00AI News Brief Hourly Summary 2026-09-17 14h : 15 posts
11:33Cognitive Extensions for Dual-Process Language Agents: Memory and Self-Reflection in Interactive Environments
11:33The Death of Schema Linking? Text-to-SQL in the Age of Well-Reasoned Language Models
11:33Compiled Agency: Frontier General-Purpose Coding Agents Build Winning Game Players from Bare Interaction – from Flappy Bird to StarCraft II and Civilization
11:33AI agent swarms are a massive waste of tokens with zero quality gain, says OpenAI Codex developer
11:33Flag Game: A Toy Model for Mechanistic Swarm Interpretability
11:33Why Infrastructure AI — Not Consumer AI — Will Define the Next Decade
11:33MUSE: Benchmarking Large Vision-Language Models on Multi-Modal Understanding in Situated Education
11:04Infinite-Parameter LLMs: Generating and Adapting Weights from Live Data
11:03Lost in Perception: Isolating Perceptual and Reasoning Failures in Multimodal Physics and Geometry Reasoning
11:03Compositional Policy Violations: When Step-Level Compliance Fails In Agentic AI Workflows
11:03Tower and NewPhotonics Ship Laser-Integrated PICs for AI Interconnect
11:03Suppressed, Not Erased: A Representational Trace of Edited Facts Survives Even Weight-Free Knowledge Editing
11:03Adecco Group rolls out Agentforce Coworker to 27,000 staff in 40-plus countries
11:03Function Lives Where Variance Doesn’t: Task-Weighted Charts of a Language Model’s Computation
11:00AI News Brief Hourly Summary 2026-09-17 13h : 15 posts
10:33Which LLM is Best for Translating Natural Language Goals to PDDL
10:33Clueing up LLMs with Tool-Augmented Deductive Reasoning
10:33CERA-MoA: Co-Evolving Routing Mechanisms with Continually Learning LLM Agents
10:32Version- and Scope-Aware Question Answering over Normative Documents: A Deployed System and an End-to-End Evaluation at Production Scale
10:32OpenRouter’s staggering token chart is the AI bubble debate in a single image

200 posts published today

Post navigation