AI news for builders and product teamsUpdated Nov 18, 2026, 07:00 UTC
AI news, straight from the source.
Releases, research, and ideas for developers and product teams. Updated every hour.
Across the feeds
Newest first
A US federal judge dismissed antitrust lawsuits from Chegg and Penske Media against Google over AI search products, ruling Google's conduct is not illegal under antitrust law. Judge Amit Mehta found no formal agreement existed between Google and the publishers.

Amazon Quick has launched Live Data in Apps, letting AI-built Quick apps query governed Quick Sight datasets in real time instead of relying on static build-time snapshots. Queries execute as the viewer, so existing row-level and column-level security rules apply, with per-dataset consent and server-side enforcement.

Google launched Guided Vision in Gemini Live on compatible Android devices, giving real-time audio descriptions of whatever the phone's camera points at, including reading small text and identifying objects. The feature is aimed at people who are blind, have low vision, or want help in specific situations.

OpenAI announced the global launch of two new ChatGPT shopping features: virtual try-on for clothing and accessories using user photos, and a Favorites library for saving products. The features use the newly launched ChatGPT Images 2.5 model, OpenAI said.

Google launched a prototype orbital compute satellite carrying a Tensor Processing Unit, part of its Project Suncatcher plan for space data centers. Google research suggests reaching about $200 per kilogram launch costs by 2035 would require SpaceX's Starship to fly roughly 1,800 missions over 10 years.

OpenAI has parted ways with three safety researchers after an internal investigation found they mishandled sensitive company information, according to a Wall Street Journal report. OpenAI confirmed the departures in a statement, saying the three violated policies on accessing and handling sensitive company information.

NVIDIA released DOCA AI agent skills on GitHub that provide verified API signatures, hardware capability requirements and build constraints for BlueField DPU development. In a 65-prompt evaluation, agents without skills satisfied 19% of checklist items versus 100% with skills.

Tavus introduced Griffin, which it calls the first "Human Interaction Model," an AI that holds real-time video calls. In a Tavus study, 48 percent of participants believed Griffin was a real person after a one-minute call, compared with two percent for previous systems.

NVIDIA's Do Inference Now (DIN) Deploy is an open-source collection of C++ samples combining ONNX Runtime with the TensorRT RTX execution provider for local AI inference on Windows and Linux. It covers speech recognition, image segmentation and image generation, with DGX Spark performance measurements included.

Databricks announced branch-based restores for Lakebase Postgres, aiming to make recovery faster in managed OLTP environments where restores have historically been slow and worsen at scale.

A Graphite study of frontier AI models found that each model has distinct writing tells, such as Claude Opus 5.5's frequent use of "dependable" and OpenAI's Astra's corrective framing. Graphite found 13,000 phrases at least twice as common in AI content as human writing.

An AWS Machine Learning Blog post shows how to implement Amazon S3 Vectors as a custom persistent memory provider for the NVIDIA NeMo Agent Toolkit (NAT), tested with NAT version 1.6. The walkthrough covers creating S3 Vectors infrastructure, writing a MemoryEditor plugin, configuring an agent workflow, and deploying on Amazon EKS.

Ideogram has announced Ideogram 4.5, a model it says edits only selected parts of an image while keeping the rest unchanged. It offers four quality tiers from 0.8 to 22 cents per image at native 2K resolution, and an open-weight release is planned.

A federal judge dismissed antitrust lawsuits from Chegg and Penske Media Corporation, which alleged Google used AI search features to divert web traffic and pressured publishers into supplying content for AI Overviews. Judge Amit Mehta ruled on Wednesday that the claims do not stand up to antitrust law.

Simon Willison released pwasm 0.2a0, a pure Python WebAssembly engine, after 42 commits made with Claude Opus 5.5. The PyPI wheel now bundles working WASM builds of MicroPython, QuickJS and Micro QuickJS, though he warns it is alpha software he would not trust.

LangChain built a model router into Open SWE's harness that cut median cost per coding task by 64% with no measurable change in quality, according to an A/B test across 973 threads. The post describes how to build a similar router.

An OpenAI News headline and excerpt argue that advanced AI may matter most for the routine work behind breakthrough ideas, and that execution could shape the next economy and the pace of progress.

Amazon Payments used a multi-objective contextual multi-armed bandit on Amazon SageMaker AI to personalize content across a product acquisition funnel. In a seven-week online A/B test, one customer population saw a high single-digit percentage relative lift in final-funnel conversion, while another saw no improvement.

Amazon Web Services' Strands Labs released Strands Decider 2B, an open-source decision model inspired by TypeSafe's Jev. It sorts between pre-decided options and reports confidence, is small enough to run locally, and briefly topped the Jevbench ranking for its size.

Shopify introduced Canvas, a site-building tool that lets merchants create and customize their online stores by chatting with Shopify's AI agent Sidekick. Changes render in real time on the store's actual code, and merchants can also edit elements directly.