US Treasury targets Chinese AI firms as power lines reshape infrastructure debates
US Treasury threatens sanctions against Chinese AI company Moonshot, while New York's record-breaking power transmission line faces operational setbacks.

Treasury Secretary Scott Bessent has accused Moonshot AI of improperly distilling Anthropic's Claude model, escalating US pressure on Chinese artificial intelligence companies. The threat of sanctions underscores deepening technological competition between Washington and Beijing, even as some US industry leaders downplay the risk.
Nvidia CEO Jensen Huang has pushed back against the narrative of Chinese AI threat, arguing that American companies maintain decisive advantages. Yet the reality is more complex: Chinese AI models are now embedded in global infrastructure and research pipelines, making simple decoupling difficult.
The sanctions pressure comes amid broader efforts to restrict Chinese access to advanced chips and AI training techniques. However, the effectiveness of these measures remains contested. Some analysts note that Chinese firms have already begun adapting their approaches, while others warn that isolation could accelerate parallel technology development in Beijing.
Grid modernization hits practical limits
While geopolitical tensions dominate tech headlines, a different infrastructure challenge is playing out in New York. A 339-mile underground power transmission line linking Quebec to Queens opened in May, becoming North America's longest underground line of its kind. The project promises to supply up to 20% of New York City's electricity demand, primarily through Quebec's abundant hydropower.
But the line has been offline for most of July. Experts voice concerns about how drought conditions upstream could affect the hydropower supply it depends on. The snags highlight a recurring problem in grid modernization: ambitious infrastructure projects often encounter operational complications once deployed at scale.
The line's performance over coming months will signal whether large transmission upgrades can reliably address US electricity demands, particularly as data centers and AI computing facilities consume ever more power. If it overcomes current reliability issues, the project could influence how policymakers approach future grid expansion.



