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 4
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 storiesPolicy
Anthropic updates usage policy to ban abusive behavior toward Claude
Anthropic has codified a new policy prohibiting sustained, needless abusive or cruel behavior toward its AI models, effective November 12.stuff.co.zaProduct
OpenAI launches Decisions API to compete with TypeSafe Jev
OpenAI released the Decisions API on Oct. 6, a service designed for fast, low-cost automated judgments that directly competes with TypeSafe's viral Jev model.fortune.comPolicy
Anthropic commits $150 million to federal Genesis Mission for AI science
Anthropic announced a $150 million, three-year commitment to the Genesis Mission, making Claude available to over 15 US federal agencies including NASA, NIH, and NSF to accelerate scientific discovery.
anthropic.comSecurity
CrowdStrike details AI tooling in South Korean bank breaches
CrowdStrike Intelligence identified infrastructure associated with a targeted campaign against South Korean financial organizations, uncovering Claude Code session histories and ARTEX configuration files that reveal the threat actor's use of agentic AI tooling.crowdstrike.comIndustry
Reflection and Mistral launch new open-weight models to challenge Chinese AI dominance
U.S.-based Reflection and France's Mistral have released new high-parameter open-weight models, Beam and Mistral Large 4, aiming to close the performance gap with leading Chinese open-source models.eu.36kr.comInnovation
Study reveals winner’s curse in self‑improving large language model loops
A new arXiv paper shows that LLMs that rewrite their own prompts can over‑estimate improvements, with evaluation bias ranging from 1 to 20 points depending on the size of the selection set.arxiv.orgIndustry
DeepSeek releases open-source tools for Huawei Ascend AI chips
DeepSeek has open-sourced DeepGEMM-Ascend and DeepEP-Ascend to lower switching costs for developers moving from Nvidia CUDA to Huawei's Ascend accelerators, though adoption remains limited by hardware access and developer skepticism.techradar.comProduct
Microsoft integrates llama.cpp and DeepSeek V4 Flash into Windows ML
Microsoft announced at its October 7 keynote that Windows ML will support llama.cpp, enabling local execution of DeepSeek V4 Flash and NVIDIA Nemotron models on Windows 11 devices with sufficient unified memory.pasqualepillitteri.itProduct
Mistral releases ML4, claiming it matches top closed AI models
Paris-based Mistral unveiled ML4, an open-weight model it claims rivals top closed systems in key areas at a fraction of the compute cost, challenging US and Chinese AI dominance.edition.cnn.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