AI News

How persistent drought is impacting fall foliage colors across the United States | thv11.com (www.thv11.com)

Scottsdale's water recycling program hits election-year backlash | National News | wboc.com (www.wboc.com)

Scottsdale's water recycling program hits election-year backlash - WGAU Radio (www.wgauradio.com)

Scottsdale's water recycling program hits election-year backlash - AP News (apnews.com)

The Year of Internal Tools (www.geocod.io)

Show HN: Air-gapped file encryption as self-decrypting HTML page (cms-sfx-demo.apeleg.com)

OpenAI Agent Bypassed Australian Medicare Portal Controls to Access Non-Public Files (thehackernews.com)

Can open-source prompt-injection detectors catch realistic AI agent attacks? (github.com)

NotchPop (www.producthunt.com)

TeamFiltration Campaign Compromises Seven Microsoft 365 Accounts Using Default Passwords (thehackernews.com)

Aurora residents gather to oppose data center near them - Denver Gazette (www.denvergazette.com)

Attackers Exploit WordPress CVE-2026-87902 Within Hours of Disclosure (thehackernews.com)

Rogue OpenAI Agent Tried to Breach Government Site in May When Prompted for Simple Data-Retrieving Tasks (slashdot.org)

Contrastive-LM Releases CLM-8B: An Open System One Model That Scores Agent Actions Up to 9× Faster Than Jev (www.marktechpost.com)

Early rogue AI agent activity and attempts to hack found on urlquery.net (transluce.org)

Six-year-old breaks women's world Rubik's Cube record [video] (www.youtube.com)

Veille quotidienne — 24/09/2026

ZooWork (www.producthunt.com)

Contrastive Language Models (contrastive-lm.notion.site)

Pulaski County passes controversial ordinance regulating data centers - thv11.com (www.thv11.com)

Faithful, Interpretable Chest X-ray Diagnosis with Artifact-free B-cos Networks (arxiv.org)

An open benchmark for machine learning-based polymer property prediction (arxiv.org)

Contrastive Learning for Authorship Verification (arxiv.org)

On the Diffusibility of High-Dimensional Latents (arxiv.org)

Intelligence Across Embodiments (arxiv.org)

MENO: Memory-Efficient Neural Operator (arxiv.org)

Discrete Diffusion Models via Evolving Variational Autoregressive Networks (arxiv.org)

EvoAudio: Recursive Self-Improvement for Audio Understanding (arxiv.org)

Evaluating ADC-only deep learning pipelines for breast cancer detection and segmentation using standalone diffusion-weighted MRI (arxiv.org)

How Sensitive Are LLM Leaderboard Claims to Hidden Model Selection? (arxiv.org)

Multitask Regression with Pairwise Fusion (arxiv.org)

Repairability of Inexact Solvers in Recursive State Estimation with Machine Learning (arxiv.org)

Benchmarking Text-to-SQL under Role-Based Access Control (arxiv.org)

Scaling Attention Head Analysis via Gradient-Based Attribution in Context-Aware Machine Translation (arxiv.org)

Field-of-View Extension in Dental Cone-Beam CT via Implicit Neural Representations and Diffusion Model-Based Refinement (arxiv.org)

Distillation for Efficient Multitask Manipulation Policies via Conditional Flow Matching (arxiv.org)

Can LLMs Catch a Rigged Backtest? A Clean-Control Calibration Benchmark (arxiv.org)

Prompt, Probe, Train, or Annotate? Single-camera sports video understanding in amateur settings (arxiv.org)

TEMPS: Temporal Sentence Embeddings for Temporal Information Retrieval (arxiv.org)

Fine-Tuning LLMs for Translation: General Forgetting Mitigation Does Not Preserve MT-Specific Instruction Following (arxiv.org)

Quantum score matching with applications to learning thermal states (arxiv.org)

ForgetMimic: Motion Unlearning for Reinforcement Learning Humanoid Control (arxiv.org)

LEAP-CBF: A Safety Filter for Uncertain Systems with Least-Effort Adversarial Potentials (arxiv.org)

PBLH Estimation from Satellite Radiances via a Dual-Encoder Transformer (arxiv.org)

Fine-Tune, Then Rectify (arxiv.org)

ASCIIBench: Evaluating Language-Model-Based Understanding of Visually-Oriented Text (arxiv.org)

Controlled Attribute-Specific Summarization of Interrogative Dialogues (arxiv.org)

Compliant with Local Controls, Collectively Discriminatory. A Governance Architecture for Multi-Agent AI in Regulated Finance (arxiv.org)

Schr\"odinger's Code Repository: Have LLMs Learned SWE-bench or Memorized It? (arxiv.org)

RelCheck: Dual-Evidence Spatial Grounding for VLM Hallucination Correction (arxiv.org)