The morning briefing.
While you slept: OpenAI's new cybersecurity model challenges rivals, Oracle cuts thousands of jobs due to AI, and NVIDIA dominates supercomputing.
AI Cybersecurity. OpenAI has launched GPT-5.5-Cyber, claiming it outperforms Anthropic's Mythos on cybersecurity benchmarks, intensifying competition. This development coincides with a Five Eyes warning that AI models specialized for cyber attacks are only months away. Japanese startup Sakana AI's Fugu also orchestrates multiple LLMs to rival top models. The global shutdown of Anthropic's Fable 5 is now creating opportunities for Chinese rivals like Zhipu GLM-5.2.
AI Workforce & Ethics. AI's impact on employment is evident, with Oracle shedding 21,000 roles over the past year, partly due to AI adoption. Ethically, Meta's Oversight Board urged better protection against sexualized deepfakes. AI is also powering the World Cup, relying on thousands of data workers for analysis. A lawsuit alleges gas stations use AI to illegally inflate prices, raising consumer protection concerns.
Global AI Infrastructure. The UK government is investing £60 million in university AI labs to develop cheaper AI systems. NVIDIA maintains its lead, powering over 400 of the world's 500 fastest supercomputers. Alibaba's chip unit T-Head tripled its capital to 1 billion yuan, signaling a major AI hardware push. Baidu unveiled its Unlimited-OCR, and WeChat is testing its Xiaowei AI assistant for 1.4 billion users.
OpenAI says new GPT-5.5-Cyber outperforms Anthropic's Mythos on cybersecurity benchmark
OpenAI has expanded its Daybreak cybersecurity initiative with the GPT-5.5-Cyber model and a partner network of over 25 security firms. The updated model focuses on automatically patching vulnerabilities, shifting from just finding them.
Five Eyes warns that AI models specialized for cyber attacks are only months away
The Five Eyes intelligence alliance warns that AI models capable of specialized cyber attacks are imminent, requiring a whole-of-organization and whole-of-society response. They emphasize the urgent need for action as AI capabilities rapidly strengthen.
Oracle sheds 21,000 roles over the past year amid wave of AI layoffs from tech giants
Oracle has reduced its workforce by 21,000 roles over the past year, stating that the adoption and deployment of AI technologies have contributed to these reductions. This reflects a broader trend of AI-driven job cuts across major tech companies.
UK government backs university AI labs with £60M to make AI cheaper
The UK government is investing up to £60 million in two new university AI labs to develop the next generation of AI systems. The funding aims to make AI cheaper and more accessible for future applications.
Sakana AI's Fugu orchestrates multiple LLMs to match Anthropic's Fable and Mythos benchmarks
Japanese AI startup Sakana AI has launched Fugu, a system that coordinates multiple AI models dynamically to compete with leading models like Anthropic's Fable 5. This approach also seeks to reduce reliance on any single AI provider.
How Anthropic’s Fable 5 shutdown could help China’s Zhipu GLM-5.2 gain ground
The global suspension of Anthropic's Claude Fable 5, due to US restrictions, is prompting developers worldwide to reassess their dependence on US technology. This situation creates an opportunity for capable Chinese rivals like Zhipu GLM-5.2 to expand their market presence.
The Oversight Board says Meta needs to do more to protect regular people from sexualized deepfakes
Meta's Oversight Board has issued several recommendations to enhance the reporting and effectiveness of protections against sexualized AI deepfakes for ordinary users. The suggestions aim to make it easier and more effective for individuals to address such harmful content.
The AI-powered World Cup runs on thousands of data workers
Human annotators in Brazil, Cambodia, and the Philippines are tracking every movement in the World Cup football tournament. This extensive human input feeds the AI systems used by teams, broadcasters, and the betting industry.
Suit Alleges That Gas Stations Use AI to Hike Gas Prices
A lawsuit alleges that gas stations are utilizing AI to unlawfully increase gas prices, potentially violating consumer protection laws. This legal action highlights concerns about algorithmic price manipulation.
NVIDIA Powers Over 400 of the World’s 500 Fastest Supercomputers
NVIDIA technology now powers 81% of the TOP500 supercomputers and 90% of the new systems on the list. Additionally, 26 systems on the TOP500 have adopted the NVIDIA Grace CPU, indicating a growing influence in high-performance computing.
Square Yards enters unicorn club with $95 Mn funding round
Real estate and mortgage platform Square Yards has achieved unicorn status after raising ₹900 crore (approximately $95 million) in a funding round. The round, anchored by EAAA Alternatives, valued the Gurugram-based company at over $1 billion.
Baidu Unveils Unlimited-OCR: Constant KV Cache Delivers SOTA Performance on Long Documents
Baidu has unveiled Unlimited-OCR, a new model that utilizes a constant KV cache to achieve state-of-the-art performance on long documents. This advancement significantly improves the processing capabilities for extensive text.
WeChat Begins Testing 'Xiaowei' Native AI Assistant for 1.4 Billion Users
WeChat has started testing its native AI assistant, "Xiaowei," for its massive user base of 1.4 billion. This move integrates advanced AI capabilities directly into one of the world's largest social platforms.
Alibaba chip unit T-Head triples capital amid AI hardware bet
Alibaba Group Holding's chip design unit T-Head has more than tripled its registered capital to 1 billion yuan ($148 million). This significant capital injection underscores Alibaba's intensified efforts to become a domestic semiconductor powerhouse amid the AI boom and China's chip self-sufficiency drive.