The Trump administration is reportedly preparing a very unusual trade move: using a Pentagon-built AI system to help set reference prices for critical minerals as part of a proposed U.S.-led metals trading bloc. According to Reuters, the plan would use the Defense Department’s OPEN program (Open Price Exploration for National Security) to price key materials in a way that attempts to strip out what U.S. officials see as Chinese market distortion.
If this sounds like a mashup of industrial policy, AI, and tariff strategy — that’s because it is.
What the administration is reportedly trying to do
Reuters reports that the idea is tied to a broader proposal, pushed publicly by Vice President JD Vance earlier this month, for the U.S. and more than 50 countries to use reference prices for critical minerals at different production stages, backed by adjustable tariffs.
The AI model would reportedly be used to generate those reference prices, with an initial focus on four minerals:
- germanium
- gallium
- antimony
- tungsten
These aren’t random picks. They’re strategically important for semiconductors, defense, and advanced manufacturing — and they’re markets where pricing can be opaque and heavily influenced by state-backed production.
Why this matters
This is more than a trade-policy gimmick. It signals a deeper shift in how Washington is thinking about economic competition with China:
- not just tariffs on finished goods,
- but active attempts to shape upstream commodity pricing.
The administration’s goal, according to Reuters and prior reporting on the OPEN/CMF effort, is to create more pricing transparency and confidence for Western miners and manufacturers so they can sign long-term supply deals without being undercut by sudden price swings or oversupply.
In plain terms: Washington wants to make it easier for non-Chinese mineral projects to look financially viable.
The logic behind the AI model
OPEN was launched by DARPA in 2023 to estimate what certain metals “should” cost after accounting for labor, processing, and other costs — while trying to factor out alleged market manipulation. Reuters has previously reported the program later moved toward nonprofit administration through the Critical Minerals Forum (CMF), which is meant to help miners, manufacturers, and investors use the data in supply negotiations.
That’s the core pitch:
- use AI to create a more “real” price signal,
- then use tariffs to defend that signal.
It’s an ambitious idea — and a controversial one.
The big problems (and why critics are skeptical)
Even if the strategy sounds clever, Reuters notes there are still major unanswered questions:
- Will the AI-derived prices be fixed or move dynamically?
- Would they apply across the whole bloc or vary by country?
- How would tariffs be applied to products that contain these minerals?
- Can the U.S. actually persuade dozens of allies to join?
There’s also a practical issue: a reference price is not the same thing as a guaranteed floor price. Former officials and analysts cited by Reuters note that competition among multiple producers could still push prices around, even behind tariff walls.
And if the policy succeeds in lifting prices, that may help miners — but it could also raise costs for downstream industries like autos and manufacturing. Reuters specifically notes antimony as an example where higher protected pricing could benefit U.S. projects while increasing costs for users.
Why this is showing up now
Reuters reports the administration is moving away from directly guaranteeing price floors for individual mining companies, partly because it lacks congressional funding for that approach. So instead of direct subsidies or guarantees, the White House appears to be exploring a market-architecture solution: AI-derived reference prices + tariffs + allied coordination.
That’s a revealing shift. It suggests the administration still wants to support domestic and allied minerals projects — but through trade design rather than direct fiscal support.
The bigger takeaway
This proposal shows how the critical-minerals fight is evolving. The next phase may not be just about permits, mines, and export controls. It may be about who gets to define the price of strategic materials in the first place.
If the plan moves forward, it could become one of the most aggressive attempts yet to use AI as a tool of industrial and geopolitical strategy — not to write emails or generate code, but to redraw the economics of a supply chain war.
Bottom line
Trump’s team is reportedly trying to turn a Pentagon AI program into a pricing engine for a China-light critical-minerals bloc. If it works, it could give Western producers more confidence to invest. If it fails, it risks adding another layer of complexity, cost, and uncertainty to already fragile supply chains.
