Tech news, September 24, 2026

In this episode of The Agent Factory, we explore the reality of building with autonomous agents alongside Ryan Lopopolo, a software engineer at Google Cloud and the person who coi…

As part of our agentic platform initiative, we are open-sourcing the Agentic Identity Broker, a central component for user-to-agent delegation, consent, and token exchange that le…

As deep learning models move rapidly from research prototypes into core enterprise infrastructure, the demand for practical, end-to-end PyTorch expertise has never been higher.

Amazon EventBridge relaunches event buses with a new enhanced Custom event bus that lets you build event-driven applications that decouple teams and can scale with your organizati…

Amazon EventBridge announces an enhanced custom event bus for organizations scaling event-driven applications across teams and accounts.

Getting to a prototype is easier than ever.

You can now require an interactive re-authentication or a multi-factor challenge before members take high-impact actions on GitHub Enterprise Cloud accounts.

What is a developer to do when they need something more tangible than a chat box?

What is spatial reasoning?

Generative AI

Hi everyone,

Expired artifacts are no longer displayed in the GitHub Actions run summary or returned by the REST API.

CoreWeave Mission Control brings frontier scale expertise to AI operations.

How Enkrypt AI's autonomous red-teaming agents adapt, chain attacks, and probe AI systems beyond what a static library can reach.

AWS Lambda durable functions are now available in the AWS European Sovereign Cloud.

LangSmith Fine-Tuning is in public beta today, and its open-source CLI, smithtune, trains on Baseten Loops.

The best model for your business will keep changing.

Microsoft will retire several standalone Azure Communication Services (ACS) offerings on September 30, 2028.

LangSmith Engine now includes Red Teaming to proactively detect agent issues and automated agent testing.

Docker's WeAreDevelopers keynote shows how Sandboxes, Kits, and Cloud Sandboxes give AI agents strong isolation and reproducible authority.

Transcribe 1 Pro is a speech-to-text model from Fish Audio tuned for interviews, meetings, and podcasts.

Trending on the Hugging Face Hub.

Ember-1 is a specialized reasoning model from Fireworks Research, built on [Kimi K3](https://openrouter.ai/moonshotai/kimi-k3).

Two compromised GitHub Actions were re-enabled with malicious tags intact, exposing thousands of downstream repositories to Mini Shai-Hulud.

Eunsoo Kim discovered that curl incorrectly handled SASL negotiation for LDAP authentication in certain circumstances.

Affects @bytebase/dbhub.

Affects org.http4s:http4s-scala-xml_2.12, org.http4s:http4s-scala-xml_2.13, org.http4s:http4s-scala-xml_3.

At OpenSSF, securing the open source software supply chain isn’t just about writing code or establishing policies.

Overview ViewSonic vCast software, which is included in ViewBoard smartboard devices, contains multiple vulnerabilities that an attacker can chained to achieve full device comprom…

Affects github.com/ixofoundation/ixo-blockchain/v8, github.com/ixofoundation/ixo-blockchain/v7, github.com/ixofoundation/ixo-blockchain/v6, github.com/ixofoundation/ixo-blockchain…

Affects decepticon-core, decepticon, decepticon-sdk.

Bulletin ID: 2026-118-AWS Scope: AWS Content Type: Important (requires attention) Publication Date: 09/24/2026 12:00 PM PDT Description: pgcollection is an open source extension t…

Top CodeQL alternatives compared on noise, fixes, cost, and more: Aikido Security, Semgrep, Snyk Code, SonarQube, Checkmarx, and Veracode Category: DevSec Tools & Comparisons

Evaluate agent responses with Jev and Arize AX.

Define and attach an outbound policy to a hosted agent in Microsoft Foundry, observe calls in Audit, and verify allowed and denied destinations under Enforced mode.

At Grafana Labs, observability is what we do.

The AI Loop is how your models constantly improve.

By Ramya Subramani and Prateek Vats In our Engineering Energizers Q&A series, we highlight the engineering minds driving innovation across Salesforce.

Over the weekend, “System One” decision models such as Jev have launched, which are...

The AWS WhisperX Deep Learning Container packages Whisper, wav2vec2 forced alignment, and speaker diarization into a GPU-ready image.

Build a multi-account architecture that keeps each team's data in its own AWS account while giving AI agents a unified way to query across them.

Learn how Aderant built an intelligent ticket triage system on Amazon Nova Lite through Amazon Bedrock, automating context gathering, classification, routing, and knowledge enrich…

Analytics teams have always modeled data for BI.

A new language-processing tool could help identify the highest-risk individuals from natural language, enabling swifter interventions.

Online sports betting company DraftKings is using AI to target customers who are most likely to place losing bets and respond to gambling promotions.

In their new book, “How AI Sees the City,” the leaders of MIT’s Senseable City Lab examine the technology’s implications for researching urban life.

npj Digital Medicine, Published online: 24 September 2026; doi:10.1038/s41746-026-03284-z Enabling equitable global health AI with privacy‑enhancing technologies

npj Digital Medicine, Published online: 24 September 2026; doi:10.1038/s41746-026-03294-x A multicenter assessment of human oversight of generative AI outputs in simulated clinica…

Key takeawaysArtificial intelligence (AI) and robots have the potential to offset the macroeconomic consequences of an ageing workforce, but the ability to automate jobs depends o…

37signals – the creator of Ruby on Rails (RoR) – moves over to agents generating nearly all its code.

Rob Bowley is “flipping tables in his head” with anger at the current media coverage of the danger of AI killing us all The risk I’m worried about isn’t a future machine deciding…

Guest Post: In science, thinking has gotten cheap but doing has not.

Human collaboration, like most things, improves with feedback.

It was the autumnal equinox a few days ago.

This is an impressive piece of malware.

CSET’s Hanna Dohmen shared her expert insight in an article published by The Wall Street Journal.

CSET’s Helen Toner was featured in an article published by Forbes.

To see what works and what breaks when agents are sent into a market, we made a miniature market of Claudes.

Asynchronous monitoring, incident investigations, and compliance audits primarily rely on agent traces to reconstruct what happened.

Latent world models are typically trained to predict factual transitions, whereas model predictive control (MPC) must compare alternative actions from the same state.

Distributed learning in embodied reinforcement-learning agents offers a degree of privacy by retaining raw sensor data on-device and transmitting only policy gradients to the serv…

Conspiratorial discourse on social media is not always expressed through explicit claims or stable lexical markers.

Coding agents have demonstrated enormous success in solving complex programming problems.

World Action Models (WAMs) couple action generation with future visual prediction for robotic manipulation.

We introduce Forte, a sensitivity type system for Rust whose soundness rests on ownership.