AI news for Saturday, 14 November 2026
The day’s AI news in one list, the biggest stories first. Each story links to its short read and to the outlets that reported it.
- AT&T plans to triple agentic AI workflows next year
A major telecommunications company is using autonomous AI agents to redesign its business processes rather than simply automating individual tasks.
Reported by SiliconANGLE
- 2 · Policy & DramaAsian governments react to Nvidia CEO Jensen Huang’s exit
Jensen Huang's departure as Nvidia CEO has prompted responses from multiple Asian governments, as the chip company plays a significant role in global AI infrastructure and has major partnerships across the region.
Reported by TechCrunch
- 3 · Policy & DramaRansomware attacks spike as criminals use AI tools
Ransomware attacks, where criminals encrypt systems and steal data to extort payments, have spiked significantly in recent months with claimed incidents rising 20 to 40 percent since spring.
Reported by Wired AI
- Vercel releases UI SDK for AI coding agents
Vercel Labs has released an open-source library of interface behaviors designed for AI coding agents to assemble into components across multiple frameworks.
Reported by MarkTechPost
- 5 · Policy & DramaElon Musk, other tech execs sign petition calling on Nvidia to reinstate Jensen Huang
A petition with more than 115,000 signatures, including support from several prominent tech executives, is calling for Nvidia to reinstate its recently ousted CEO.
Reported by TechCrunch
- Cmpnd releases dspy-wsp weight-space optimizer
Weight-space programming is a technique where a frontier language model reads and directly edits another model's weights to improve its performance, without using gradients, training loops, or fine-tuning datasets.
Reported by MarkTechPost
- 7 · Policy & DramaHelen Toner says AI safety needs more than trusting leaders
Helen Toner, a board member who voted to fire Sam Altman at OpenAI, argues that relying on the trustworthiness of individual AI leaders is not a viable safety strategy.
Reported by The Verge
- Stanford releases Memorilla memory system for frozen LLMs
Memorilla is an open-source memory system that compresses entire document collections into just 16 memory tokens for use with frozen large language models, allowing the models to access information without expanding their context windows.
Reported by MarkTechPost
- Cognigy releases AI agent software focused on production reliability
A company released a new version of its AI agent software designed for contact centers.
Reported by SiliconANGLE
- 10 · Agents & ProductsShield helps technicians hand off legacy portal work with OpenAI
An IT services platform built AI agents that can navigate older computer systems and portals that lack modern API connections, allowing them to complete technical tasks that previously required a human technician to manually log in and…
Reported by OpenAI
Every story, by topic
One short email a day with the stories that matter. Free. Unsubscribe with one click.