Updated daily488 verified stories
Models News tracker.
New frontier, open-weight, multimodal, reasoning, and specialist AI models—plus the evidence behind their capability claims.
- 38stories this week
- 488sources cited
- 488verified timeline entries
News archive
AI news archive — Page 21
Source-checked AI stories, newest first, for people who need to understand AI without chasing hype.
Verified sourcing
Every story links to the strongest available evidence: original sources when available, otherwise clearly attributed reporting.
Plain English
What happened, why it matters, and what to watch — without the jargon.
No filler
When the signal is thin, we publish nothing rather than padding the feed.
More stories
9 storiesProduct
Cognition releases Fusion harness for lower-cost AI agent performance
GIGAZINE reports that Cognition released Fusion, a harness for GPT-6 Astra and Claude Fable that matched competing agent systems in benchmarks while reducing reported costs by up to 39%.gigazine.netInnovation
Werewolf benchmark finds LLMs can overvalue trusted accusers
A benchmark of 40 open-weight LLM configurations found that accusations can shift models’ beliefs even when the accuser is aligned with the opposing side.arxiv.orgPolicy
China offers BRICS an open-weight AI base while export rules remain unclear
Trending Topics reports that China proposed a shared BRICS AI ecosystem while possible export controls could limit access to its strongest open-weight models.trendingtopics.euIndustry
Loft Orbital and Marlan Space finalize $1 billion AI satellite deal
SatNews reports that Loft Orbital and Marlan Space finalized a $1 billion agreement for 50 LEO satellites using Mistral AI models for onboard data analysis.satnews.comSecurity
Anthropic says Claude-enabled attacks are spreading across cybercrime
Anthropic says threat actors used Claude in disrupted cyber, espionage, fraud, surveillance, weapons and biological-misuse operations between December 2025 and August 2026.
anthropic.comInnovation
Bodhan AI and NVIDIA release open models for Indian languages
EdexLive reports that IIT Madras’ Bodhan AI and NVIDIA have launched four open-weight models for Indian-language speech, translation and document processing, with education and public digital infrastructure as target uses.edexlive.comSecurity
Der Spiegel reports Houthi attempt to use Claude for missile software
Der Spiegel reports that a Houthi-linked cell in northern Yemen tried to use Anthropic’s Claude Code to develop guidance, navigation and control software for aircraft, while Anthropic says its safeguards blocked many but not all requests.spiegel.deIndustry
Moonshot says it filed a police report over Kimi K3 rumors
BeInCrypto reports that Moonshot AI called online claims about its founder and employees fabricated, said it had filed a police report, and faces unverified allegations about routing some Kimi users’ requests to Anthropic’s Claude.beincrypto.comSecurity
Anthropic report details Iranian and Houthi use of Claude for weapons
Anthropic's September 2026 threat report reveals that Iran-linked actors and Houthi rebels used Claude AI to target U.S. naval forces, develop hypersonic missile guidance systems, and build mass surveillance tools.tomshardware.com
One useful briefing each week
Keep up with AI without living in the feed.
Get the week’s verified AI news, original data, useful tools, learning picks, and fresh AI jobs.
Reach people who are learning AI
Hiring an AI professional or launching a useful AI product? Put it in front of people who came here to learn and act.
Post an AI jobSubmit an AI tool