The morning briefing.
While you slept: Samsung deploys ChatGPT Enterprise, China blocks Claude Fable 5, and NVIDIA cools AI servers with a 45°C breakthrough.
Global AI Adoption. Samsung is deploying ChatGPT Enterprise and Codex to all employees in South Korea and its Device eXperience division worldwide. This move signifies a major enterprise adoption of advanced AI tools for internal operations. Meanwhile, Tencent is testing an AI assistant within WeChat, China's most popular app, aiming to integrate AI services into daily consumer life. These developments highlight the accelerating global trend of integrating AI into both corporate workflows and mass-market applications.
AI Geopolitics & Workforce. Geopolitical tensions are increasingly shaping the AI landscape, as seen with the US government's order blocking non-US access to Anthropic's Claude Fable 5. This action underscores the strategic importance of domestic AI models for countries like China, which are actively cultivating their own capabilities. Concurrently, educational systems are adapting, with Chinese universities cutting language majors to prioritize AI and robotics degrees. London businesses also report that half their workforce lacks necessary AI skills, prompting increased training investments.
Frontier Models & Accountability. Innovation in frontier models continues, with Sakana AI's Fugu orchestrating multiple LLMs to compete with benchmarks set by Anthropic's Fable and Mythos. This approach aims to reduce reliance on single AI providers and enhance model performance. However, concerns persist regarding AI accountability; research suggests LLMs not only reflect biases but actively police them, raising questions about ethical deployment. The challenge for self-driving cars, for instance, lies not just in their functionality but in proving how their opaque AI systems make decisions.
Hardware & Emerging Tech. The foundational hardware supporting AI continues to evolve rapidly. NVIDIA has achieved a 45°C breakthrough in liquid cooling for its AI servers, significantly improving energy efficiency for large-scale AI factories. In the memory sector, YMTC's NAND market share has climbed to 13%, intensifying global competition in critical components. The on-device AI battle is also heating up among smartphone giants like Xiaomi and Huawei, each pursuing distinct strategies for mobile AI integration.
Samsung rolls out ChatGPT Enterprise and Codex to employees in South Korea
Samsung Electronics is deploying ChatGPT Enterprise and Codex to all employees in South Korea and its Device eXperience division worldwide. This initiative aims to integrate advanced AI tools into the company's global operations.
Tencent tests AI assistant in China's most popular app, WeChat
Tencent is testing an AI assistant within WeChat, China's most popular app, to leverage its vast user base and expand the adoption of its AI services. WeChat is an indispensable part of daily life in China.
Chinese universities are cutting language majors to make way for AI
Chinese universities are dropping programs in translation and foreign languages. This shift is to accommodate new degrees in embodied intelligence, AI, and robotics, reflecting a national strategic pivot.
Half of London's businesses say workforce not equipped for AI age
Half of London's businesses report their workforce lacks the necessary AI and digital skills to meet organizational requirements. Companies are planning to increase training investments to address this skills gap.
Lithuanian startup Superpal raises €500K for AI coworker platform built inside Slack
Vilnius-based AI startup Superpal has raised €500,000 to develop its platform, which functions as a fully autonomous AI coworker within a company’s Slack environment. FIRSTPICK led the funding round.
Europe is doubling down on its own AI infrastructure, calling it essential
European leaders emphasize the importance of owning and controlling the entire AI stack to build regional AI muscles. This strategy aims to reduce reliance on external technologies and foster local innovation.
LLMs do not merely reflect the bias of their training, they police it
New research suggests that large language models (LLMs) do more than just reflect biases present in their training data. They actively police and enforce certain biases within their outputs.
Self-driving cars aren’t the challenge – proving how they think is
The primary challenge for self-driving cars is not merely their functionality but establishing accountability for their opaque AI systems. This requires proving how these systems make critical real-world decisions.
Sakana AI's Fugu orchestrates multiple LLMs to match Anthropic's benchmarks
Japanese AI startup Sakana AI is launching Fugu, a system that coordinates multiple AI models on the fly. Fugu aims to compete with leading models like Anthropic's Fable and Mythos, reducing dependence on single AI providers.
Claude Fable 5 Blocked for Non-US Users: Domestic AI Is the Only Safe Bet for China
The US government's order blocking foreign access to Anthropic's Claude Fable 5 signals a new era of AI export control. This reinforces the strategic importance of China's domestically developed AI models.
Xiaomi vs Huawei On-Device AI: Decoding the AI Strategies of 8 Major Smartphone Giants
The on-device AI battle among major smartphone makers is intensifying, with Xiaomi and Huawei leading with distinct approaches to mobile AI. This report decodes the AI strategies of eight major smartphone giants.
Boruikang Files for IPO as China's First Brain-Computer Interface Public Company
Boruikang, known for the world's first approved invasive BCI medical device, has filed for a STAR Market IPO. The company aims to fund further R&D and commercialization of its brain-computer interface technology.
YMTC NAND market share climbs to 13% as global competition intensifies
Yangtze Memory Technologies Corporation (YMTC) has increased its global NAND flash memory market share to 13%, up from 8% in 2025. This growth intensifies global competition among memory chipmakers.
NVIDIA's 45°C Breakthrough to Cool AI’s Biggest Machines
NVIDIA's newest AI servers can run their cooling liquid up to 45 degrees Celsius, a breakthrough that significantly improves energy efficiency. This higher temperature limit is crucial for cooling large-scale AI factories.