This page is available in English: Home, Bio, Work, Blog, Contact, Privacy, Photo rights and the ticker. Some images may still contain Dutch text.
Archive From the archive of The Latest A.I. News
This is an archive page with the A.I. news of Monday 19 October 2026. This day was compiled afterwards from reports that appeared at the time, so it was not kept on the day itself; a lot may have changed since. Today's news is on The Latest A.I. News.
Compiled afterwards on
Microsoft and Amazon focus on AI agents and chips, while a hack at OpenCode shows how exposed API keys are. Nvidia releases new open models.
The new features, including Guardian Agents, supervise the agents a company uses, detect up to 60 agent types and limit their access. Risky actions need human approval. Microsoft's security chief says frontier models are unpredictable and not built to follow organizational norms.
With a one-line configuration change, PyTorch apps run on Amazon's Trainium instead of Nvidia hardware. Amazon says about 85 percent of AI apps use PyTorch and is contributing changes to the framework to loosen its ties to specific chips. That could give developers more choice but also extra configuration work.
Attackers broke into an older system and affected 1.4 percent of users of the OpenCode Console, where some store their API keys. The stolen keys led to unauthorized charges, sometimes above $1,000. Anthropic and Google blocked the keys and OpenAI helped with refunds; the app and command-line tool were not affected.
The valuation is more than double that of late 2024. The inference-focused chip startup, built around open-weight models, stayed independent after Intel and Qualcomm reportedly showed interest. The money will scale chips for open models and video generation.
Nemotron 3.5 Super VL has 120 billion parameters and, Nvidia says, finishes agent tasks faster than rival models from Anthropic and OpenAI. Two smaller Bolt models run locally on new AI PCs and DGX Spark workstations; the larger is available now, the smallest later this month.
An announcement is reportedly close, but talks are ongoing and terms could change. It would follow an earlier deal of the same size about a month ago. Both companies are pursuing IPOs, and Nscale reportedly faces pushback on its proposed $30 billion valuation.
New York's mayor urged residents not to accept AI as inevitable and to demand accountability from AI companies. He pointed to US policy that mostly allows self-regulation, and to city measures such as a ban on AI in elementary and middle schools and a proposed disclosure rule for AI-generated property listing images.
Next up: An announcement of the second Nscale-Anthropic deal appears close.
Every day has its own page. They stay up for three months.
Everything here is a summary of reports from other media, not original research. The label (Certain, High, Medium, Low) is my estimate of how reliable the source is. The summaries are generated automatically with AI and may contain mistakes; read the article itself for the details (the star marks the leading source). Sources: leading A.I. and tech sites and the official blogs of the A.I. companies.
