Asian and US tech stocks slump on AI financing and capacity fears
Global tech markets suffered a sharp decline as investors questioned the return on AI investments and feared memory chip overcapacity. South Korea’s Kospi index has lost nearly 40 percent of its value since its June peak, while US indices show signs of strain.
Asian stock markets experienced a sharp decline on Thursday, with South Korea hit hardest due to its heavy exposure to AI data center expansion. Key indices saw losses of 10 to 12 percent, driven by rising costs for CoreWeave's debt and a Korean analyst report warning of memory chip overcapacity by 2029. The Kospi index has lost nearly 40 percent of its value since its June peak, a drawdown comparable to previous major financial crises. While European and US markets did not immediately follow the decline, South Korean and Japanese markets showed signs of recovery by Friday.
In the United States, the S&P 500 was off 1 percent from its record set the previous week, while the tech-heavy Nasdaq was down more than 3 percent from its peak. A Goldman Sachs basket of AI stocks lost 20 percent in five trading days. Within the Nasdaq 100, 55 of 100 companies are more than 20 percent below their 52-week highs, and seven have halved, including AppLovin, Rocket Lab, Intuit, Ansys, Western Digital, CoreWeave, and Strategy. The index remains resilient because Apple is only 1.5 percent below its all-time high and Microsoft is 4 percent below.
US and German bond yields hit multi-decade highs on inflation fears
US 10-year Treasury yields reached 5.365 percent, the highest level since 2002, while German Bund yields rose to 3.65 percent. The rise is attributed to inflationary pressures from tariffs, geopolitical conflicts, and AI-driven demand for components.
Meta releases MuseSpark Open with first-party Apple hardware support
Meta launched MuseSpark Open, an 816 billion parameter model optimized for Apple M5 Ultra chips. Nvidia and Black Forest Labs also released new open-weight models for enterprise agents and image generation.
- Microsoft releases MAI Base 2 with superior pre-training efficiency
- Anthropic reports $16.8 billion Q3 revenue and targets $110 billion annualized
- A16Z partner argues AI coding success does not guarantee operational reliability
- Anthropic grants METR researchers employee-level access for safety audits
- Anthropic tests AI self-replication across home network devices
Eight states sue OpenAI over ChatGPT safety and data practices
Eight U.S. states filed lawsuits against OpenAI, alleging ChatGPT poses risks to minors and that personal data was collected without consent. The states seek financial penalties, age verification, and a halt to self-improving AI development.
- Musk denies Starshield role in Ukraine service disruption
- Anthropic grants third-party evaluators employee-level access
- White House task force sets agenda for AI safety report
- Ondo prepares U.S. launch following SEC innovation exemption
- Russia earns 1.1 trillion from fossil fuel exports despite sanctions
Researchers define five pillars for high-quality agentic context
A new framework from DX, Capital One, GitHub, University of Victoria, and Google identifies context quality as the primary bottleneck in AI agent performance. The authors argue that treating context as a core engineering discipline is essential to prevent compounding failures in autonomous workflows.
Gartner predicts 80% of AI cloud providers will fail by 2030
Gartner forecasts that most specialized AI cloud providers will not survive independently by 2030 due to high costs and rapid hardware depreciation. Market stress is already evident as CoreWeave bonds trade at distress levels and its stock falls sharply.
AI decompilation threatens SaaS and gaming business models
Developers are using AI to reconstruct proprietary software, causing stock drops and forcing companies to pivot. Analysts warn that AI-driven code generation is undermining traditional IP and hardware tethers.
YouTube Music and Deezer introduce AI content labels
YouTube Music is rolling out a labeling system to identify generative AI in music, while Deezer launches a filter to exclude such tracks. These moves follow similar disclosure initiatives by Apple Music and Spotify.
Character AI declares insolvency after Google investment runs dry
Character AI has declared insolvency after exhausting a 2.7 billion dollar investment from Google. The collapse highlights the difficulty of establishing a viable revenue model for AI companion platforms despite active market demand.
Circle launches ARK blockchain with permissioned validators
Circle has deployed its ARK network, a permissioned blockchain designed for institutional financial contracts and AI-driven agents. The launch integrates with Circle's cross-chain transfer protocol to support deterministic settlement and broad interoperability.
FBI arrests OpenAI engineer for planting listening device
An OpenAI engineer was arrested for installing a listening device in a San Francisco conference room. The company used its own AI model to help identify the employee, highlighting growing concerns over corporate espionage.