The recent, dramatic fallout between the U.S. Department of Defence and Anthropic—the creator of the Claude AI model—transcends the boundaries of a domestic American policy dispute.
This seismic event compels policymakers and business leaders across Asia to reassess the nature of technological power. For nations lacking the resources to build their own “frontier” AI stacks, the position of a mere consumer has become increasingly precarious.
The Illusion of Corporate Control
At the heart of the Anthropic-Pentagon clash lies a fundamental question regarding whether AI companies can truly prevent their technologies from being used for purposes they deem unethical. Anthropic attempted to enforce “red lines” by specifically prohibiting the use of Claude for mass domestic surveillance and fully autonomous lethal weapons systems.
However, the reality of Operation Epic Fury suggests that corporate control is often a fragile illusion. Despite the U.S. government officially blacklisting Anthropic as a “supply chain risk” on February 27, reports indicate that military agencies continued using Claude to help plan and execute strikes in Iran for days afterwards.
The Pentagon’s demand for an “all lawful use” standard—requiring that models be free from vendor-imposed ethical constraints—demonstrates that once a technology is integrated into state machinery, the owner may lose the power to pull the “kill switch.”
For Asian business leaders, this raises a terrifying prospect: proprietary intellectual property could be “commandeered” by state power under the guise of national security, regardless of any established terms of service.
A Global Governance Vacuum
The dispute also exposes a dangerous gap in international oversight. Anthropic is a Public Benefit Corporation, which carries a statutory obligation to prioritise AI safety over shareholder returns. Yet, the U.S. executive branch dismissed these ethical safeguards as “woke” obstructionism.
Currently, no comprehensive international framework governs the value systems of “Big AI” tech owners. We are witnessing a period of “regulation by contract,” where the ethical architecture of humanity’s most consequential technology is decided in opaque negotiations between CEOs and military secretaries.
If the “safest” company in the world is blacklisted for its principles while rivals swoop in to offer “unrestricted” models, the market is incentivising a race to the bottom. In this environment, the most ethically lax providers gain the most influence.
The Asian Response: From Monitoring to Sovereignty
Regional powers are already shifting their strategies in response to this volatility.
● India is viewing the Anthropic-DoD dispute as proof that the “American Stack” is a strategic liability. At the recent India AI Impact Summit, officials signalled a pivot toward “Digital Public Infrastructure” for AI. They are prioritising indigenous models that remain unaligned with foreign military mandates.
● Singapore has initiated an Economic Strategy Review via its Smart Nation Office. As a global hub, the city-state fears that U.S. export controls will force local firms to choose between the U.S. military market and the global commercial sector. Consequently, Singapore is doubling down on “Model Diversification” to ensure public services do not suffer from a foreign-driven service blackout.
● Malaysia is closely monitoring the risks to its digital infrastructure as it rolls out the National Education Plan 2026–2035. The National AI Office (NAIO) has raised concerns regarding the “ethics of export,” fearing that AI tools in Malaysian schools could be subject to surveillance features demanded by foreign governments.
This has shifted Malaysia’s focus toward “Auditable Open Source” and domestic hosting.
Especially for Malaysia, who have just issued the Call for Tender to develop Malaysia’s AI Governance Bill, this discussion must be on the minds of policymakers and not just “follow” and adapt, especially the EU Policies.
A Call for Regional Strategy
Asian policymakers must recognise that AI neutrality is a myth. The Anthropic-Pentagon feud proves that AI models are not just tools; they are geopolitical instruments. We cannot rely on the “restraint” of superpowers to protect the interests of our citizens.
The path forward for Asia must involve heavy investment in Sovereign AI. Reducing dependency on foreign frontier models that can be “weaponised” or restricted by foreign executive decrees is no longer optional.
Furthermore, we must move beyond “regulation by contract” toward binding multilateral treaties on autonomous weapons and surveillance.
The era of “move fast and break things” has ended. In its place is a high stakes struggle for the hearts and minds of people beyond borders. For Asia, the choice is clear: work together, lead the conversation on ethical governance, while at the same time, be on top of the evolution of AI to find our own AI spaces through innovation.
Otherwise, we will risk becoming spectators and casualties of someone else’s war. – March 29, 2026
K V Soon is a digital transformation specialist and reader of SCOOP