Skip to main content
Geopolitics & DefenseGovernment & Policy

US, Allies Fortify AI Supply Chain Defenses Against Hostile Actors

Officials from various countries examine digital infrastructure at a secure shipping terminal.

"The security of our peoples cannot be separated from the security of our economies and those of our closest partners," the San Francisco declaration says — a simple line that aims to recast chips, models and ports as national-security issues to be managed jointly.

San Francisco declaration: who signed and when

The San Francisco declaration was signed on November 15 and was jointly released by the US administration and the governments of Australia, India, Israel, Japan, South Korea, Qatar, Singapore, the United Arab Emirates, and the United Kingdom. The move, described in the declaration as a shift from shared principles to "a working partnership with shared responsibilities," explicitly ties economic and national-security protection to a defined set of technologies and infrastructure.

Pax Silica: turning a year's rhetoric into operational cooperation

The declaration formalises work under the Pax Silica initiative, first announced in December last year. Pax Silica frames its remit around "vital parts of the global technology supply chain," including frontier foundation models, compute and semiconductors, advanced manufacturing, minerals refining and processing, and energy. The initiative’s stated purpose is collective stewardship of those elements — though observers have noted the list of participants does not include China, a fact that has fuelled commentary characterising Pax Silica as an effort to "China‑proof" AI and chip supply chains.

What the partners say they will protect: technologies and infrastructure named

The declaration enumerates what partners must defend: "software, materials, knowledge, components and inputs…essential to AI supply chains." It specifies protection where "theft, illicit transfer, or acquisition by hostile actors would confer a meaningful military or intelligence advantage, or undue economic leverage over the participants, or whose unauthorized transfer or misuse could pose significant risks to the economic and national security of the participants."

Beyond chips and models, the text singles out critical infrastructure — ports, terminals, transport, communications and logistics networks — and the ICT and energy systems these depend upon. The implication is clear: security for AI supply chains extends from silicon and code to the physical chokepoints that move them.

Jensen Huang's resignation and the Megaspeed probe: a proximate pressure point

The declaration arrives amid heightened political pressure to tighten export controls, a dynamic sharpened by the recent resignation of Nvidia's CEO Jensen Huang. According to the reporting, Huang admitted to an affair with a woman suspected by US authorities of being a spy for Beijing. She worked for Megaspeed, a Singapore‑based company that bought Nvidia hardware and was suspected of making it available to Chinese users.

US, Singaporean, and Malaysian police investigations are continuing to probe whether chips were diverted or accessed remotely by Chinese firms. Those inquiries have added momentum to lawmakers — from both Republican and Democrat parties, according to the reporting — who want further measures to prevent China from accessing advanced chips used for AI processing.

What this means for policymakers, enterprises, and adversaries

  • Policymakers and regulators: The declaration reinforces a push toward coordinated export controls and shared responsibility for supply-chain security; the reporting notes bipartisan interest in additional restrictions on advanced chips.
  • Affected enterprises and procurement leaders: Companies that design chips, make the machines that fabricate them, or move them through global logistics chains should expect a heightened regulatory and operational focus on where know‑how and hardware flow.
  • Adversaries and threat actors: A recent report claimed US export controls are not stopping AI labs in China from using restricted Nvidia chips, with many cutting‑edge models reportedly trained on Blackwell GPUs — a data point that complicates the declaration's aim of denying sensitive capabilities to hostile actors.

Diplomacy, precedent, and the next test

Advocates of Pax Silica stress the novelty of its shared-responsibility framing. "What makes the San Francisco Declaration historic is that Pax Silica partners have agreed to treat the security of those technologies and the infrastructure behind them as a shared responsibility," Jacob Helberg, the US under‑secretary of state for economic affairs, told the Financial Times.

Yet the declaration lands while high‑level diplomacy is in motion: it is understood that President Trump is to meet China's premier Xi Jinping this week on the sidelines of the APEC summit in Shenzhen. The declaration also sits in a longer arc of American policy: the reporting notes efforts to control Chinese access to US technology began under the first Trump administration and continued under President Joe Biden's leadership.

That chronology matters because the declaration's effectiveness will be tested not only by policy coordination among signatories but by whether export controls and investigations can close the gaps flagged by recent reporting and the Megaspeed inquiry. The partners have put shared language and an explicit list of targets on the table; the practical question left open by the facts in the public record is whether shared words will translate into shared, enforceable results.

Original story