The US government is investigating Chinese artificial intelligence developer Moonshot AI over allegations that its Kimi K3 model improperly used outputs from Anthropic's Fable model and was trained with restricted Nvidia GB300 chips accessed outside China.
For the past three years, large language model (LLM) development has moved along the same track: models have gotten bigger, parameter counts have climbed, reasoning has improved, AI agents have matured, and long-context windows have kept breaking records. The industry's debate has also centered on model capability, computing scale, and how far the field is from artificial general intelligence (AGI).
Nvidia CEO Jensen Huang's whirlwind visit to Japan in mid-July extended well beyond product announcements and public appearances. Alongside a joint event with gaming giant Sega Corporation, Huang used the trip to promote what he has repeatedly described as Nvidia's next major AI frontier—physical AI—sparking renewed discussion across Japan's industrial sector about the country's role in the emerging technology.
Volkswagen, BMW, and Mercedes-Benz are reshaping their growth strategies as the global auto market enters a low-growth era, with profitability under pressure and EV demand slowing. Instead of chasing volume, the three German automakers are focusing on multi-powertrain platforms, deeper AI integration, and cross-border partnerships.
