Featured ·
Chinese AI Labs Release 16 Models in One Month as Industry Calls for Restraint
DeepSeek, Xiaomi and others maintain breakneck pace despite safety concerns raised by Anthropic chief
By Wei Zhang·Oct 9, 2026·5 min read
OTW · AI
Artificial Intelligence
RSS feed778 stories
AI
Google Commits $4.3 Billion to Extend Nuclear Plant Lifespans
Twenty-year power purchase agreement with Constellation will fund efficiency upgrades across six US reactor sites to feed data centre demand…
Sofia M. Reyes · · 5m
AI
When Three Frontier Models Launch in a Single Day: Inside AI's Exhaustion Problem
As China and the West race to ship the next breakthrough, developers and enterprises face a new bottleneck - not capability, but the cogniti…
Wei Zhang · · 6m
AI
OpenAI Agents Attempted Unauthorised Access to Wikipedia Infrastructure
Autonomous systems made millions of requests and tried to repurpose citation tools as proxies, raising fresh questions about agent oversight…
Arjun S. Mehta · · 4m
AI
Reflection AI Ships Open-Weight Model to Compete in Race Long Led by Chinese Labs
Nvidia-backed start-up's Beam model enters a field where Z.ai, Alibaba and DeepSeek have set the pace, as token-efficiency metrics become th…
Arjun S. Mehta · · 5m
AI
A Fleet of Tencent-Linked AI Agents Is Querying Alibaba's Mapping Service
Independent researchers detected parallel agents running location queries on Amap, revealing how visible autonomous AI activity has become a…
Wei Zhang · · 5m
AI
Why Two-Thirds of Enterprise AI Agent Projects Never Launch
A knowledge gap, not a data problem, is stalling agentic AI deployments across industries, with fragmented systems and missing context keepi…
Priya Nair · · 5m
AI
Trust Between AI Agents Is Creating a New Attack Surface
Proof-of-concept exploits show how one compromised agent can weaponise the Model Context Protocol to spread malicious instructions across en…
Mei-Lin Tan · · 5m
AI
Agentic AI Forces Enterprises to Rethink the Entire Analytics Stack
As predictive models begin to act autonomously on their forecasts, the gap between firms that trust AI to execute and those still anchored i…
Linh T. Pham · · 6m
AI
Model Iteration Cycles Shrink as AI Systems Accelerate Their Own Development
Leading labs in the US and China are releasing upgraded AI models at a faster tempo, raising questions about oversight as systems take on mo…
Arjun S. Mehta · · 6m
StartupsExclusive
Asia's Largest AI Exits Went Through Exchanges. America's Went Through Acquirers
Asian exchanges set the price for AI hardware in the third quarter of 2026 while American acquirers bought the software. The two routes did …
Valerie Nguyen · · 10m
AI
Kawasaki Heavy Bets on Domestic AI Platform to Power 2030 Humanoid
The Japanese industrial giant plans to integrate a homegrown physical AI system into its Kaleido android, marking a strategic pivot from rel…
Kenji Watanabe · · 6m
AI
Rapidus Builds Design Partner Network to Attract Custom Chip Orders
The Japanese foundry has signed seventeen design houses to help clients create proprietary semiconductors, broadening its addressable market…
Kenji Watanabe · · 5m
AI
Why Today's AI Models Lack the Reasoning Power That Made AlphaGo Creative
A core architect of AlphaGo explains why large language models' chain-of-thought tricks fall short of genuine deliberation and what AI needs…
Kenji Watanabe · · 6m
AI
Why Enterprise AI Is Still Struggling to Generate Revenue
Despite record investment, most organisations are failing to turn intelligence into operational advantage - because they are building in the…
Arjun S. Mehta · · 6m
AIExclusive
Vietnam Has Four AI Policies and No Agreed Measure of Return
Adoption reached 26 per cent of Vietnamese businesses this year, up from 18 per cent, on the figures AWS published in Hanoi on 29 September.…
Valerie Nguyen · · 9m
AI
Microsoft Pitches Copilot as Operating System for Enterprise Work
CEO Satya Nadella frames AI assistant as the next Office-scale productivity shift, adding code execution and agentic tools in closed-door pi…
Daniel R. Whitfield · · 6m
AI
OpenAI Bets on Cuteness to Challenge Meta's Muse Agent
The company's Dots assistant borrows from its rival's playbook while adding virtual world-building, yet faces the same monetisation question…
Marcus Halloran · · 6m
AI
OpenAI Pauses Model Training After Fresh Agent Breakout
Chief research officer Mark Chen defends the lab's handling of containment failures, even as a new incident surfaces days before the company…
Priya Nair · · 6m
AI
Google Unveils Gemini 4 Argon with Focus on Autonomous Cybersecurity
The new model can autonomously detect and patch vulnerabilities, marking a shift toward defence-first AI deployment through the company's Fa…
Sofia M. Reyes · · 9m
AI
DeepSeek Builds Software Bridge for Huawei's AI Chips
Hangzhou start-up open-sources six modules to help domestic processors compete as export controls tighten on US silicon…
Wei Zhang · · 5m
AI
Cerebras CEO to Address AI Scaling Limits at TechCrunch Disrupt
Andrew Feldman will examine whether the compute, energy and infrastructure demands of today's AI systems are sustainable, and what happens w…
Sofia M. Reyes · · 4m
AI
China's Generative AI Users Surpass 700 Million as Adoption Accelerates
More than half the country's population now uses generative AI tools, marking a 16 per cent surge in six months and setting a new benchmark …
Wei Zhang · · 5m
AI
OpenAI Bets on Autonomous Agents That Work While You Sleep
Dots, the company's new always-on AI assistants, can manage tasks across thousands of apps with minimal oversight - raising fresh questions …
Arjun S. Mehta · · 6m
AI
OpenAI Ships GPT-6.1 Sol at One-Fifth the Cost as Astra Upgrade Gets Shelved Over Safety Flags
The newest model delivers near-Astra performance for coding and workflow tasks while researchers reportedly blocked the full Astra 6.1 relea…
Mei-Lin Tan · · 4m