Tech news, September 25, 2026

You can now use an in-product validator for enterprise managed settings for GitHub Copilot.

CMake Tools 1.24 includes several reliability fixes and productivity improvements.

The Unicode CLDR 49 Beta is now available for specification review and integration testing.

Gartner has recognized AWS as a Leader in the 2026 Gartner Magic Quadrant for Container Management for the fourth consecutive year.

The enterprise and organization repository-level Copilot usage metrics reports now break down how long pull requests spend in each stage of review.

Amazon Transcribe now lets you encrypt your custom vocabularies, custom vocabulary filters, and custom language models at rest with a customer-managed AWS KMS key that you own and…

Private saved views for repository issues Repository issues pages now support private saved views, making it easier to create and save personalized views of your issues.

Pulumi Neo works on infrastructure the way an engineer does: it clones repositories, edits files, installs dependencies, runs previews, and produces intermediate work along the wa…

EDB’s Ata Ul Haseeb Chaudhry is a senior director of Treasury and Financial Operations who manages multiple teams across multiple time zones.

Queries for workflow runs in the GitHub Actions API and UI now return a less precise but more accurate count of records when you search by workflow, event, status, branch,… The po…

Azure HorizonDB is a fully managed, PostgreSQL-compatible, cloud-native database service designed for scalable, high-performance workloads.Azure HorizonDB, now in public preview,…

AI coding agents are making it easier for people to start writing Rust, but is the code they’re writing good?

Windows App Development CLI v0.7.0 has shipped, and we’re excited to share it.

Agentic autofix now uses Copilot Memory for customers who’ve enabled it.

Welcome Sheila Vashee!

Compute (EC2) opened the front door to the first generation cloud.

The new AG-UI.NET SDK enables.NET services to expose and consume interoperable agent endpoints using Microsoft.Extensions.AI.

A milestone that belongs to our customers.

A thank you message for the community that made this possible.

This week’s releases add new models to Copilot, local sandboxing in the Copilot app, and updates to Copilot in Slack, Microsoft Teams, JetBrains, and VS Code.

Seed Audio 1.0 is ByteDance Seed's non-streaming audio generation model.

Jev Router picks the best model and reasoning effort for each request, balancing quality, speed, and cost.

Kev 4B is a small open-weight decision model from Jared Palmer, built as a LoRA adapter and pointer head on Qwen3.5-4B-Base and served over the same /v1/systemone contract as Type…

Perceptron Mk1.5 is Perceptron's embodied reasoning model for physical agents.

Trending on the Hugging Face Hub.

Gemini 3.5 Transcribe is a speech-to-text model from Google.

See how a global automotive technology manufacturer used NodeZero to reduce testing overhead, cut comparable test runtime by 50%, and scale security validation across more than 23…

Unit 42 security experts address critical cybersecurity misconceptions, offering practical insights to help your organization reinforce its enterprise defenses.

Posted by Haitam Lazaar on Sep 25 Hello oss-security, An untrusted search path vulnerability leading to Local Privilege Escalation (LPE) was identified in GNU libextractor for ver…

Posted by David M.

A U.S.

Posted by David M.

Posted by David M.

Posted by David M.

Socket is joining the OpenJS Security Stewardship Program to fund Node.js vulnerability research, maintainer remediation, and security releases.

IntroductionIn August 2026, Zscaler ThreatLabz analyzed a phishing campaign that used fraudulent Google ads to target Ledger hardware wallet users.

Add hybrid search to a Postgres app with Pinecone: full-text and vector search in one index, Postgres as the source of truth.

Learn how Accenture and Snowflake use AI and a unified data platform to accelerate M&A due diligence, de-risk integration and drive deal value.

With Codex, GPT-Live-1, and GPT-6 Astra, Proaction builds, operates, and sells modern fleet management faster.

This post is for platform and SRE teams running 24/7 network operations who want AI Agents and Model Context Protocol (MCP) for intelligent, automated event response, and for indi…

If you manage Enterprise Support for a large AWS organization, you've likely been asked which accounts a single Support line item comes from and how it was calculated.

Describe the interface you need in plain English, then let the agent build a live surface you can both use and update—so you spend less time adapting to tools and more time gettin…

Data is the foundation of AWS Cloud migration, and choosing the right storage services is one of the earliest, most critical decisions enterprises face.

A completed storage assessment looks simple: an AWS Storage service recommended per workload, capacity and performance recommendations, and cost.

Data Protection and disaster recovery is an important part of any migration.

Learn how to scale Mixture-of-Experts (MoE) reinforcement learning on Amazon EKS using Elastic Fabric Adapter (EFA) and DeepEP.

The PKG Center for Social Impact expands Code.Tulsa experiential learning program.

This post introduces a delegation model that gives architects a structured way to assess whether agentic systems meet their authorization and access control requirements.

npj Digital Medicine, Published online: 25 September 2026; doi:10.1038/s41746-026-03248-3 Large language models as judges for clinical generative AI evaluation

npj Digital Medicine, Published online: 25 September 2026; doi:10.1038/s41746-026-03104-4 LEME: open large language models for ophthalmology with advanced reasoning and clinical v…

npj Digital Medicine, Published online: 25 September 2026; doi:10.1038/s41746-026-03279-w Artificial intelligence-mediated clinical communication between providers and patients or…

In 2023 most people doubted that there could be more than 1 or 2 frontier model labs.

I feel like someone who reads this blog will want to go to this: Families are invited to dive into the fascinating world of marine biology during an exciting, hands-on Family Squi…

1,000+ facilities across 60+ operators mapped, built retail-first and flipped by AI, largest hyperscaler leases 1/5 national capacity, 100MW in 12 months, Eastern Data Western Com…

If you’re anything like me, you’re sick of hearing about so-called “AI”.

Earlier this month, Anthropic published a long report detailing all of the Claude misuses it detected.

Exactly on this day, September 25 2001, Apple shipped version 10.1 of their OS X operating system.

CSET’s Sam Bresnick joined ABC News for an interview discussing U.S.-China competition in artificial intelligence.

Representation autoencoders (RAEs) reuse features from a pretrained visual encoder as reconstruction and diffusion latents, integrating strong visual representations into image ge…

Reasoning models often generate very long reasoning traces, making inference computationally expensive.

We give a gap-free differentially private algorithm for the principal component analysis (PCA) problem with Gaussian data.

Recent literature has shown a strong connection between optimization and sampling.

Generative AI systems are increasingly used, but aligning their outputs with user requirements poses a continuing challenge.

Large language models (LLMs) implicitly infer attributes of their users and adapt their behavior accordingly, yet these beliefs remain difficult to inspect and causally manipulate.

Low-rank adapters (LoRA) make it cheap to fine-tune a large language model once per task, but combining several independently trained adapters into one model remains difficult: me…

3D scene graphs provide a structured representation of complex environments by encoding objects, their semantic attributes, and the spatial and functional relationships between th…

The Manufacturer Usage Description (MUD) standard allows IoT manufacturers to define expected network behaviors in a MUD file.