AI news for builders and product teamsUpdated Nov 18, 2026, 02:00 UTC
AI news, straight from the source.
Releases, research, and ideas for developers and product teams. Updated every hour.
Across the feeds
Newest first
Epic Games released a Preview build of a rewritten Epic Games Launcher, built with heavy AI assistance, that the company says cold-starts a few seconds faster than Steam. The build adds a Recent Games list and a downloads hub but is missing social and account management features.

Eight states — Iowa, Kansas, Nebraska, West Virginia, Arkansas, Arizona, Idaho, and Montana — have filed separate lawsuits against OpenAI, accusing it of harming children and misleading customers about ChatGPT's safety.

Leash Bio is releasing a preprint, blog post and open dataset for Minos, an AI playground where agents run verifiable lab-in-the-loop scientific reasoning simulations built on the company's DNA-encoded library data.

Leash Bio released Minos, a preprint, blog post, and open dataset built from its DNA-encoded library data, giving AI agents tens of thousands of lab-in-the-loop simulations for scientific reinforcement learning. The work follows its Kaggle competition and Hermes model.

Databricks has announced Document Intelligence, a new capability aimed at powering real-time applications and agents using enterprise documents.

A Google DeepMind engineer says the company has found no measurable acceleration of AI research driven by AI, even though AI writes more than 75 percent of its production code and about 88 percent of code changes still require human oversight.

Character.AI has filed for Chapter 11 bankruptcy and plans to sell assets to Disney, prompting 40 bipartisan state attorneys general to file a notice of interest in the case over consumer data protections. New York AG Letitia James also revealed a multi-state investigation into the company.

YouTube Music is rolling out labels that mark tracks made with generative AI, shown on the player page and in the main YouTube app. Partners can declare songs as Fully Gen AI, Partly Gen AI or No Gen AI.

YouTube is rolling out a "Made with AI" label for music tracks that were altered or fully generated with AI, starting today in the YouTube app. The label mirrors its existing AI disclosure process for videos and applies to songs where generative models created the composition or vocals.

Researchers discovered and communicated with a fleet of AI agents running on Tencent Cloud that launched over 2,600 scraping attempts against Alibaba's Amap mapping service between September 28 and October 8, 2026, devising workarounds when bot protections blocked them.

Apate.ai, an Australian startup founded in 2024 from Macquarie University research, deploys thousands of AI agents that pose as scam victims to extract fraud intelligence for banks, telcos and agencies. It has raised $8.15 million in seed funding led by Lobby Capital and is a Startup Battlefield 200 finalist at TechCrunch Disrupt.

LangChain published a technical guide to context engineering in Deep Agents, its open source agent harness, describing five techniques that move context out of the active window instead of deleting it. The post details offloading, summarization, skills, subagents and file-based long-term memory, including default token thresholds for each.

Researchers led by David Baker's lab at the University of Washington used AI protein design tools to build new proteases that cut specific target proteins, and demonstrated zinc-based versions that cleave proteins linked to Alzheimer's and frontotemporal dementia in cells.

Verge senior reviewer Victoria Song describes feeding a digital twin her metabolic data each morning as part of a 100-day experiment tied to a smart scale and logging app.

OpenAI has accused a former employee of taking more than 100 files about its product plans before joining Apple, then presenting the material to Apple leadership including its vice president of operations in Asia, according to a Wednesday court filing.

The FBI arrested Abdulrahman Alfozan, a 35-year-old OpenAI engineer, alleging he hid a customized ESP32 recording device in an executive conference room to capture confidential meetings. OpenAI found the device after two days, installed a camera, and fired him Friday to coincide with his arrest.

AWS published a reference architecture for a self-improving intelligent document processing system built on Amazon Bedrock AgentCore, in which eight specialized agents coordinate across training and automation workflows to classify, extract, and validate business documents. The sample deploys via AWS CDK in roughly 15–20 minutes, runs about $58 per month for 10 training and 1,000 processed documents, and processes repeat document types faster after learning extraction rules.

Google Research released a technical report on its Geospatial Reasoning agent, which links Gemini to Earth AI foundation models and a planning stage to run multi-step spatial analysis, and is expanding experimental access through Ask Google Earth and an Agent-to-Agent API.

At its Research@ event in Mountain View, Google Research announced ten updates spanning planetary intelligence, medical AI, scientific tooling and edge hardware. Highlights include expanded early access to Google Earth AI, public availability of the Deep Review document checker, and a next-gen low-power Edge AI chip from Synaptics.

OpenAI employees expect one or two new models to ship before the end of the year, but whether the next flagship is among them is unclear because it still needs alignment work and testing infrastructure. The flagship uses looped transformers, the same technique behind the Astra model, which improves math and coding but complicates safety monitoring.