329 articles
AI technology breakthroughs, research papers, patents, and technical innovations from leading AI startups.

Anthropic publishes training method to suppress agentic misalignment in AI agents, highlighting limits of chat-based RLHF Anthropic has published a new training technique designed to reduce "agentic m...
Alibaba's Qwen3.7-Max ranks 5th globally and 1st in China on Artificial Analysis leaderboard Alibaba's latest flagship foundation model, Qwen3.7-Max, has achieved the position globally and in China...
DeepSeek forms Harness code agent engineering team, signaling pivot from research lab to product company DeepSeek has confirmed it is building a new team called "Harness" focused on code agent enginee...

Google I/O 2026: Gemini Omni Flash Bridges Reasoning and Creation for Video Generation At Google I/O 2026, Google DeepMind introduced Gemini Omni Flash, the first model in the Omni family that can gen...
NetEase Youdao (网易有道) open-sources Confucius 4 foundation model with chain-of-thought compression for cost reduction NetEase Youdao has released the 4.0 version of its vertical education foundation mo...
Alibaba Cloud upgrades its full-stack Agent technology system, unveils in-house AI chip Zhenwu M890 Alibaba Cloud at a summit on May 20 announced a sweeping, Agent-first re-architecture of its entire...

Google CEO Sundar Pichai and Google DeepMind leadership unveiled the company's most significant product refresh in 25 years at the I/O 2026 developer conference, centering on deeply embedding AI agent...

Ant Group launches new AI model and open-source strategy at Tech Day 2025 Ant Group, the Chinese fintech affiliate of Alibaba, unveiled a new proprietary AI foundation model and announced an open-sour...

LetinAR raises $18.5M for AI glasses optics, targets 2027 IPO South Korean startup LetinAR has raised $18.5 million from Korea Development Bank and Lotte Ventures, among others, ahead of a planned 202...
**Weilan Tech's BabyAlpha A3 robot dog bypasses Nvidia compute stack with custom heterogeneous cluster** Chinese robotics company Weilan Technology (蔚蓝科技) has unveiled the BabyAlpha A3, a consumer-gra...

SAP Blocks Open-Source AI Agents, Protecting Licensing Model and Proprietary AI Assistant Joule German enterprise software giant SAP has begun actively blocking unauthorized open-source AI agents—spec...

OpenAI GPT-5.5 Tops AI Safety Institute Cybersecurity Evaluation, Outpaces Anthropic Mitos in Cybersecurity Evaluation OpenAI's latest generative AI model, GPT-5.5, has achieved the highest score in a...
Zhipu AI (智谱) releases open-source flagship model GLM-5.1 with 8-hour sustained task execution Chinese AI lab Zhipu AI (智谱) has released GLM-5.1, its newest open-source flagship foundation model, whic...

Kled and Luel dispute reveals defensibility challenge for AI data marketplaces Kled founder Avi Patel has accused YC-backed Luel of copying his company's approach to paying individuals for personal da...

Pixal3D turns single image into GLB mesh, advancing cheaper 3D asset production TencentARC's Pixal3D, a SIGGRAPH 2026 project from Tsinghua University, Tencent ARC Lab, and Victoria University of Well...

AI exploit benchmarks put Mythos at the center of cyber startup strategy. New testing from Berkeley RDI's ExploitGym benchmark shows Anthropic's Claude Mythos Preview succeeded on 157 of 898 real-worl...

OpenClaw updates AI Agent platform, signaling paradigm shift to agentic systems OpenClaw, described as the world's first consumer-facing AI agent platform, has released a major update in March 2026. T...

Microsoft has revoked Anthropic's Claude Code licenses for thousands of internal developers, designers, and project managers across its Experiences + Devices organization, transitioning them to its ow...

Greg Brockman Takes Control of OpenAI’s Products in Executive Shake-Up OpenAI has made official the promotion of co-founder Greg Brockman to lead product strategy, consolidating ChatGPT, the coding ag...
Anthropic is shifting focus to compete with OpenAI and Microsoft over the agent control plane, the operating layer of AI agents, rather than just model quality for enterprise adoption. Why it matters:...