The global Pax Silica alliance, initiated by the US to protect AI supply chains, has expanded its influence to 24 countries. Simultaneously, the US administration has demonstrated unprecedented intervention in the commercial model release cycle, requiring OpenAI to limit public access to the new GPT-5.6.

image
image

What Happened

The Pax Silica alliance, which now includes the EU, Germany, the Netherlands, and Greece, plans to include Kazakhstan, Argentina, Chile, Costa Rica, and Panama. The initiative covers the full cycle: from critical mineral mining to chip manufacturing and data center infrastructure. In parallel, OpenAI has agreed to a phased release of the GPT-5.6 model under government supervision, limiting access to a narrow circle of approved companies. Meanwhile, Microsoft introduced an update to Copilot in Excel, adding support for custom Markdown-based Skills and connectors to financial data from PitchBook and FactSet.

Context

Pax Silica is forming secure technological supply chains to reduce dependence on China. This is occurring against a backdrop of regulators shifting from soft oversight to preemptive AI lifecycle management (safety-driven gatekeeping), where access to frontier models is becoming an object of state control.

Why It Matters for the Industry

The industry is entering a phase of intense geopolitical and regulatory fragmentation. The creation of closed equipment supply chains and the standardization of safety protocols through government restrictions may slow the pace of innovation, but will ensure predictability for the infrastructure layer. A 'gated release' standard is being established for frontier model developers.

Why It Matters for Users

For professionals in the financial sector, automation tools are becoming deeper through the integration of specialized data into LLM workflows. However, for the broader range of developers and startups, the GPT-5.6 precedent creates barriers to accessing SOTA solutions and increases uncertainty in planning the implementation of new models due to regulatory risks.

What Is Not Yet Known / Limitations

There are differing assessments of the consequences: while new integration tools increase predictability for the corporate sector, access restrictions to new models could become a critical obstacle for solo developers and startups.

Sources

Author

Look at AI, Editorial Team