Former Hugging Face Asia-Pacific ecosystem head Tiezhen Wang commented on diverging AI release strategies: Chinese companies favor open-weight models while US firms like OpenAI and Anthropic keep models closed. He argued that distillation is a neutral technique, noting the irony that US companies train on public internet data yet try to prevent others from reusing that knowledge. Wang stated all AI-generated content should be free of copyright to avoid power abuse by compute-rich entities. He observed that China's plentiful open-weight models lead to drastically lower token costs, prompting Chinese internet companies to encourage maximum token usage, push employees to become AI-native developers, and even ban manual document writing.
The First Proof project tested four AI systems on ten original, unpublished research-level math problems created by mathematicians for this purpose. All problems were never included in any model's training data, and solutions were scored by anonymous expert reviewers from relevant fields. The AI responses showed frequent hallucinations and a critical absence of literature citations, failing to reference any sources. The evaluation confirmed that current reasoning models cannot yet match top human mathematicians. This was the first assessment to simultaneously satisfy three key standards: frontier math problems, no training data leakage, and expert human evaluation.
Despite recent announcements from Nvidia, SpaceX, Google, and startup Starcloud about building orbital data center constellations with AI GPUs, a closer look at the physics reveals major challenges. The claimed benefit of free cooling in space is a misconception: in a vacuum, only radiative cooling works, requiring huge radiator surfaces to prevent chip overheating. Solar power requires complex sun-tracking systems, and cosmic rays degrade panels, radiators, and chips. Space maintenance is extremely difficult, necessitating redundant systems, and a rough cost comparison shows running AI GPUs in space costs at least an order of magnitude more than on Earth. Orbital data centers may have niche uses but are currently economically unviable.
On June 10, 2026, the Munich Regional Court ruled that Google's AI Overviews are the company's own content, not conventional search results, making Google directly liable for any false statements. The case was brought by two Munich publishers after AI Overviews falsely linked them to scams and subscription traps, and Google failed to adequately respond to cease-and-desist letters. The court found that the AI rewrites and judges information in its own language and structure, and that the generated content often contradicts the linked sources, thus representing Google's own assertions. Because Google developed and controls the AI and its algorithms, it holds ownership of the output, and traditional search engine liability rules do not apply to AI-generated summaries. This ruling establishes that platform operators can be held responsible for harmful AI-generated text that damages third parties.
Apple announced at WWDC 2026 that its new Siri AI is powered by Google Gemini, with processing on-device or via private cloud. The assistant uses personal context to search files, emails, and photos, performs cross-app tasks, answers on-screen questions, and retrieves web knowledge. A dedicated Siri app syncs conversation history across devices through iCloud. The service is not available in the EU due to privacy regulations and is also pending approval in mainland China.
OpenAI has confidentially submitted an S-1 registration to go public, allowing preparations without public financial disclosure. The company has not determined the listing date or share count, stating it will choose the optimal timing. In its most recent funding round in March 2026, OpenAI reportedly raised $122 billion at an $852 billion valuation, though its valuation now trails Anthropic. Together with SpaceX and Anthropic, OpenAI is one of the most anticipated IPOs, with potential combined market caps of $4 trillion.