keynews.ai

Key AI news stories in enterprise and tech.

For practitioners and enthusiasts

#StoryPillarKeyRankSource
1Claude Opus 5.5, GPT-6 Sol, GPT-6 Luna, and a new price warFrontier88Simon Willison
2Introducing GPT-6 Sol and LunaFrontier88OpenAI Blog
3Anthropic’s biolab made a discovery it’s comparing to CrisprFrontier78The Verge AI
4Anthropic Releases Claude Opus 5.5: Fable 5.1-Level Performance at 40% Lower Running Cost Than Opus 5Frontier78MarkTechPost
5Anthropic and OpenAI roll out cheaper models in first release since call for slowdownFrontier78CNBC Technology
6Xiaomi's affordable flagship AI leads the open models, and Anthropic says Claude helped get it thereFrontier78The Decoder
7GPT-6 Astra and Claude Fable turn robot arms into slapstick killer robots in new safety benchmarkRisingFrontier75The Decoder
8SpaceX launches Grok 4.7 with long-horizon processing, safety upgradesFrontier75SiliconAngle
9Deepmind was built to chase AGI, but its new chief just wants Gemini 4 out the doorFrontier72The Decoder
10Anthropic says Claude discovered a new enzyme system, but CRISPR researchers call it routine genome miningFrontier72The Decoder
11Black Forest Labs launches FLUX 3 Action, an open robotics AI modelFrontier72The Decoder
12Google launches two benchmark-topping speech generation modelsFrontier72SiliconAngle
13NVIDIA Releases Nemotron 3 Diarization: A 100M-Parameter Open-Weight Model That Tracks 8 Speakers in Real TimeFrontier72MarkTechPost
14Introducing MentalHealthBenchFrontier72OpenAI Blog
15Anthropic unveils Opus 5.5: powerful performance, still premium priceFrontier72AI Business
16Alibaba launches Qwen Audio 3.1 with new models and slashes AI audio prices by up to 95 percentFrontier72The Decoder
17Alibaba Unveils Zhenwu V900 — and Plans Qwen Models With Up to 10 Trillion ParametersFrontier72TechRepublic
18Kyutai Releases Voice of Reason: A Speech-Native Model that Solves Spoken Math with Reinforcement LearningFrontier72MarkTechPost
19Meta admits Muse’s likeness to OpenClaw isn’t a coincidenceFrontier72TechCrunch AI
20Google’s robotics unit Intrinsic open-sources its foundational infrastructure for intelligent robotsFrontier72SiliconAngle
21OpenAI says its internal model solved over 100 long-standing math problems after just a month of trainingFrontier72The Decoder
22Anthropic Builds Biology Lab to Test What Claude Can Do in the Real WorldFrontier72TechRepublic
23Jev introduces a new shape of LLM - System One, aka Decision ModelsFrontier72Simon Willison
24OpenAI forms math advisory group as its AI resolves more than 100 open problemsFrontier72TechCrunch AI
25Alibaba Qwen Releases Qwen-Image-2.1: A 7B Open-Weight Model for Image Generation and EditingFrontier72MarkTechPost
26StepFun Launches Step 5 Preview: A 600B-Total, 27B-Active MoE Model With 1M Context for Long-Horizon Agentic WorkFrontier72MarkTechPost
27Tether addresses AI underinvestment in Africa with open-source machine translation modelsFrontier72CIO
28Anthropic is operating a lab that conducts biology experimentsFrontier72TechCrunch AI
29PrismML Releases Ternary Bonsai 2 27B: A 5.9 GB Apache 2.0 Model Retaining 98.2% of Qwen3.8 27B PerformanceFrontier72MarkTechPost
30Jina AI Releases jina-ocr-v1: A 3.4B MoE Document Parser With Built-In Speculative Decoding for Low-Budget GPUsFrontier72MarkTechPost
31PrismML launches Bonsai 2 27B, a high-intelligence AI model so small it fits on consumer hardwareFrontier72SiliconAngle
32The U.S. says China's AI progress is down to 'distillation.' But is it that clear cut?Frontier72CNBC Technology
33Alibaba Qwen Releases Qwen3.8-Omni-Flash: A 1M-Context Omni-Modal Model Built Around Agentic Audio-Video Understanding and Tool UseFrontier72MarkTechPost
34Qwen3.8-Omni-Flash undercuts Google's Gemini Flash pricing while matching its multimodal benchmarksRisingFrontier72The Decoder
35Alibaba Qwen Team Releases Qwen3.8-LiveTranslate: A Real-Time Interpretation Model That Cuts Average Lag to 2.3 Seconds Across 60 LanguagesRisingFrontier68MarkTechPost
36Top AI experts badly underestimated how fast the field is moving, study findsFrontier68The Decoder
37Simulated students that make realistic mistakes help AI tutors learn fasterRisingFrontier62The Decoder
38Vals, backed by Andreessen Horowitz, is looking to become the gold standard for AI benchmarkingRisingFrontier65TechCrunch AI
39OpenAI Expands Outside Safety Reviews Into Model TrainingFrontier62TechRepublic
40Anthropic engineer explains why Claude's writing got worse although the model got smarterFrontier62The Decoder
41How UK AISI and EvalEval Are Making Benchmark Results ReproducibleFrontier62Hugging Face Blog
42AI Models Built From Rat Brains Just Got Closer to RealityFrontier62Wired AI
43Linkup Research Releases SPARSEUP: A 149M-Parameter Open-Source Sparse Embedding ModelFrontier62MarkTechPost
44Anthropic brings in Accenture for AI safety testingFrontier62Financial Times Technology
45World model companies are keeping a lot of secretsFrontier62TechCrunch AI
46Anthropic wants you to know Claude leads a quarter of its research, but "lead" doesn't mean what you thinkFrontier62The Decoder
47Visible chains of thought are a safety advantage for AI, but that transparency is slipping awayFrontier62The Decoder
48Anthropic and OpenAI need truly independent safety evaluators, experts say in public letterFrontier62CNBC Technology
49OpenAI Introduces Triage Framework and Case Studies to Report Model MisalignmentFrontier62InfoQ AI/ML
50Tencent's Gander aims to keep talking while it works in the backgroundFrontier57The Decoder
51BottleCap AI Releases ThinkingCap-Qwen3.8-27B: 37.2% Fewer Thinking Tokens at a 0.86pp Accuracy CostFrontier52MarkTechPost
52Compressing Streaming Neural Audio Encoders via Latent-Space DistillationFrontier52Apple Machine Learning
53Priorities and principles for effective third party assessmentsFrontier52OpenAI Blog
54Advancing AI for biology: Teaching models to design and characterize antibodiesFrontier52Amazon Science
55tokenizers v1: encode, decode and scaling, measuredFrontier52Hugging Face Blog
56Pruning LLMs Like a Physicist: Block Removal as an Ising Optimization ProblemFrontier52Hugging Face Blog
57How to Guide Your Language FlowFrontier48Apple Machine Learning
58Accelerating vision-language models with LFM2.5-VL-DSparkFrontier45Hugging Face Blog