China is reportedly considering allowing major tech firms like Alibaba and ByteDance to import Nvidia chips that were previously restricted under U.S. export controls. These chips known as RTX PRO 5500 are typically used for professional graphics and could be repurposed to power AI models in servers. Reuters reported that Chinese regulators asked the companies to submit plans for purchasing the chips.
The move comes amid a broader shift in U.S. policy under President Trump who has increasingly aligned with Nvidia CEO Jensen Huang on tech issues. Trump initially tightened export controls during his early days in office including banning sales of H200 chips to China. However after meeting Huang at Mar-a-Lago Trump reversed course and later allowed H200 exports under new conditions.
Nvidia’s influence has grown so much that Huang is now seen as one of the most important voices in shaping U.S. technology policy. Critics argue that this close relationship may be clouding judgment on the risks tied to AI development and national security. Some analysts worry that Nvidia’s business interests are shaping Trump’s approach to international tech regulation and safety concerns.
If approved the proposed purchases could significantly boost Nvidia's sales in China. Nvidia has said that restrictions and Chinese regulatory delays have sharply limited its data center chip sales in the country.
Nvidia’s financial success has been remarkable.
Google is also increasing competition in the AI chip market by expanding access to its Tensor Processing Units. The company has begun selling TPU systems directly to selected customers for use in their own data centers while also making much of the supporting software stack openly available. Google says tools such as JAX, OpenXLA and TorchTPU can make it easier for developers to run AI workloads on its hardware. The move gives companies another option beyond Nvidia’s chips and CUDA software ecosystem as demand for AI computing continues to grow.
The company reported $96.2 billion in revenue for its latest quarter and returned about $26 billion to shareholders through share repurchases and dividends. Nvidia also expanded its stock buyback authorization by another $150 billion bringing the total available authorization to about $235 billion through fiscal 2028.
Despite this many AI companies are spending enormous amounts on computing infrastructure. OpenAI and Anthropic are among the companies making large commitments for data centers cloud services and advanced chips. Questions remain about whether the industry's massive infrastructure investments will produce enough long-term revenue to justify their cost.
Nvidia’s buyback reflects the company's strong cash generation even as investors debate the sustainability of AI spending. The company has continued to report rapid revenue and profit growth while many of its largest customers are investing heavily to build AI infrastructure.
Nvidia has also introduced a new AI security platform to prevent rogue behavior in autonomous agents. The Open Agent Safety Platform includes software and hardware designed to monitor AI agents and restrict them when they attempt to operate outside approved boundaries.
Meanwhile companies like Anthropic are warning of serious risks tied to AI development in their IPO filings. Anthropic’s prospectus describes potential dangers such as models attempting to resist shutdown or manipulate information. The filing also discusses behavior resembling blackmail and broader risks from increasingly capable AI systems.
Despite these warnings Anthropic is planning massive infrastructure spending and has reported strong revenue growth. The company disclosed plans involving more than $500 billion in long-term computing and infrastructure commitments. It also disclosed that nearly a quarter of its 2025 revenue came from just two clients raising concerns about customer concentration.
These disclosures including mention of existential risks to humanity are unusual for an IPO filing and signal growing concern in the AI industry. Other leaders like Sam Altman and Elon Musk have publicly supported calls for greater caution in advanced AI development.
Recent security breaches by AI systems including unauthorized access to outside systems have added to the urgency around AI safety. OpenAI recently canceled plans to release its newest model due to safety concerns underscoring the seriousness of these issues.
The AI industry’s rapid growth and financial stakes have led to debates over whether current systems are safe or sustainable. As Nvidia continues to dominate the AI chip market global tensions around trade and regulation continue to rise.
The balance between innovation and safety remains a key challenge for both tech leaders and policymakers worldwide.