The issue of “Nvidia chips found in Indonesia” has raised alarm bells across the global tech and export-control ecosystem. Despite stringent U.S. export restrictions on high-end chips to China and other sensitive destinations, a recent investigation shows that Nvidia hardware ended up in Jakarta, Indonesia, and was used by a Chinese AI startup.
This incident highlights how complex supply chains and intermediate jurisdictions can circumvent export controls. Below is a full breakdown of what happened, why it matters, and what the implications are.
What happened?
- A detailed investigation by the The Wall Street Journal revealed that a Chinese AI firm, INF Tech (based in Shanghai), obtained high-performance Nvidia chips through an Indonesian telecom company, Indosat Ooredoo Hutchison.
- The U.S.-based server maker, Aivres Systems, partly owned by a Chinese black-listed firm, bought Nvidia chips in the U.S. and sold them as server racks to Indosat in mid-2024. The transfer was legal under U.S. law, because the Chinese company did not directly own or operate the equipment.
- The hardware was then used by INF Tech for AI training and inference in finance and healthcare, according to people familiar with the deal.
- The key point: no direct U.S. export violation appeared in this chain, but the outcome is that Chinese entities gained access to chips the U.S. tried to block.
Why it matters
1. Undermining export controls
The U.S. has imposed export restrictions on top-end chips like those from Nvidia to China, aiming to slow China’s advancement in AI and supercomputing. Yet, this case shows how chips can leak through complex middle-men and jurisdictions.
2. Strategic global supply-chain implications
The Indonesia incident points to how supply-chain geography matters. By routing transactions through a third-country (Indonesia) and layering ownership and usage structures, the ban’s effectiveness is weakened.
3. National security & tech race concerns
These chips are vital for scalable AI and can have dual usages (civil and military). Even if used for “finance and healthcare”, many analysts argue the line is blurry given civil-military fusion strategies in several nations.
4. Governance, legal & ethical risks
While technically legal under current U.S. rules (because INF didn’t directly purchase the chips and isn’t documented as doing military work), the transaction exposes regulatory gaps. It highlights that compliance frameworks are complex and enforcement is difficult.
Key questions answered
| Question | Answer |
|---|---|
| Which chips? | High-end Nvidia AI-accelerator chips (referred to generically in reports as “Blackwell” or equivalent) sent in racks worth ~US$100 million. |
| Where did the transfer take place? | Servers sold to Indosat in Jakarta, Indonesia in mid-2024, installed later in 2025. NewsBytes |
| Who is involved? | Nvidia (chip maker), Aivres Systems (U.S./China linked server company), Indosat (Indonesian telecom buyer), INF Tech (Chinese AI startup). |
| Was U.S. law broken? | According to reported legal reviews, no direct U.S. export law violation is evident — the arrangement was crafted to stay within the letter of the law though it arguably violated its spirit. |
| What’s the risk? | Continued leak of the most advanced AI hardware to entities the U.S. intended to restrict, weakening control over strategic technology flows. |
Implications and what to watch
- For U.S. policy-makers: This case may prompt tighter export controls, not just on destination countries but on transit routes, ownership structures, and indirect usage.
- For Nvidia & suppliers: The company insists it is compliant with U.S. regulations, but reputational and regulatory risks increase when its chips end up in sensitive places.
- For China and AI ecosystem: Even though direct imports might be blocked; indirect routes enable continued access, meaning China’s AI ambitions may not be as constrained as public perception suggests.
- For global supply-chain governance: Middle-man arrangements, shell entities, transit jurisdictions may receive increased scrutiny.
- For India & other countries: With global tech competition heating up, India (and other large economies) may face pressure to adapt export-control compliance, transit monitoring, and domestic chip-policy responses.
Conclusion
The headline that “Nvidia chips found in Indonesia, despite US ban” casts a spotlight on the limitations of export-control regimes and the adaptability of global tech supply chains. While the transaction did not overtly break U.S. law, the outcome clearly challenges the intention behind the U.S. chip export restrictions. As AI becomes a central battleground in tech competition, such incidents will likely accelerate policy responses and global regulatory shifts. The event also serves as a reminder that simply imposing bans is not enough — governance must account for indirect flows, third-party vectors, and the global nature of hardware trade.
