Full Summary
This Wednesday morning, Google's highly anticipated Gemini 3.5 Pro AI model faces a delay, now expected to launch in July. Both TradingView and TipRanks confirm this pushback, citing Google's need for further adjustments and integration of feedback from early testers. Meanwhile, the US government is actively seeking voluntary review agreements for advanced AI models. Reuters, citing a New York Times report, indicates Meta is the only major US developer yet to sign on, though Crypto Briefing notes Meta's stated intention to do so. This framework, stemming from a June 2026 executive order, allows federal agencies up to 30 days of confidential access to frontier AI models before public release. In China, the AI race is heating up. The Times of India reports Zhipu CEO Tang Jie challenging Elon Musk's timeline for Chinese frontier AI, suggesting a rival to Anthropic's Fable 5 is much closer than anticipated. Zhipu's GLM-5.2 model, an open-source 744-billion-parameter model, already shows benchmark performance close to Anthropic’s Opus and reportedly surpasses OpenAI’s GPT-5.5. On the medical front, the South China Morning Post announces a Chinese surgical robot, "Toumai Remote," has received EU approval, enabling remote laparoscopic surgeries. NC AI is set to launch VARCO 3D 2.0 in July, a next-generation 3D generative AI that significantly improves shape accuracy and can generate 3D assets in just three minutes, as reported by The Korea Times and 디지털투데이. This could drastically cut costs and time in various industries. Finally, NVIDIA is expanding into life sciences with agentic AI, aiming to automate complex scientific workflows in drug discovery, according to koreabiomed.com. This could fundamentally change how scientific software is used. These rapid advancements and governmental oversight efforts mean the AI tools you use, from search engines to medical diagnostics, are under constant scrutiny and evolution, directly impacting their reliability and the data privacy you can expect.