Global competition to build AI chips and physical infrastructure for models and autonomous systems

Global Competition Accelerates in AI Chip Development and Infrastructure for Autonomous Systems

The race to dominate the future of artificial intelligence is intensifying across the globe, with startups and established companies alike vying to develop advanced AI chips and the necessary physical infrastructure to support increasingly autonomous and model-driven systems. This surge reflects not only technological ambition but also strategic geopolitical moves, as nations recognize the critical importance of semiconductor capacity in shaping economic and military power.

Rapid Growth of AI Chip Startups and Private Investment

Across Korea, Japan, and other regions, a flurry of activity is centered on building the hardware backbone for AI applications:

• Startups like Boss Semiconductor and BOS Semiconductors are raising significant funds—Boss secured ₩87 billion to scale mobility-focused AI chips, while BOS raised $60.2 million in Series A to advance AI chip development. These investments underscore a focus on creating specialized processors for autonomous vehicles and other mobility applications.

• MatX, a startup challenging industry giants like Nvidia, recently secured $500 million in Series B funding aimed at developing next-generation AI processors. Their goal is to accelerate hardware for both civilian and military use, reflecting a broader trend of militarized AI hardware development.

• Additionally, private firms such as Wayve in the UK have attracted $1.5 billion to innovate in autonomous vehicle and drone technologies, inherently tied to military reconnaissance and autonomous systems.

Strategic National Programs and Semiconductor Investments

Beyond startups, national initiatives are emphasizing the importance of semiconductor capacity as a strategic asset:

• Japan’s Rapidus, a government-backed venture, recently received a $1.7 billion boost, aiming to secure domestic supply chains for advanced AI chips. This investment highlights Japan’s strategic move to counter Chinese dominance in semiconductor manufacturing and to bolster regional autonomy in critical technology sectors.

• Similarly, Rapidus’s substantial funding underscores a broader effort within Japan and Asia to develop sovereign chip capabilities, essential for supporting autonomous military and civilian systems.

• Saudi Arabia has also entered the fray with a $100 billion fund dedicated to AI, semiconductors, and advanced tech sectors, signaling a desire to diversify its economy and achieve technological independence amid global competition.

The Geopolitical and Security Dimension

This surge in hardware development is tightly intertwined with the broader context of international competition in AI. Countries are racing to build not just models but the physical infrastructure—AI chips, secure data centers, and autonomous hardware—that enables militarized AI applications.

• The strategic importance of these chips is underscored by ongoing concerns about model theft, reverse engineering, and illicit proliferation. Allegations against Chinese firms like DeepSeek illustrate how adversaries are attempting to siphon advanced AI capabilities, potentially embedding proprietary features into autonomous military systems—raising risks of escalation and destabilization.

• The proliferation of large language models (LLMs) such as Qwen3.5 and multimodal systems like V4 has democratized access to powerful AI tools, increasing the risk that militarized AI could spread beyond borders. The diffusion of these models enhances capabilities for surveillance, autonomous weapons, and reconnaissance—further intensifying the global arms race.

Industry and Policy Responses

In response, global leaders and companies are investing heavily in tools to detect and prevent illicit activities:

• Google, among others, is developing advanced detection and attribution technologies to identify reverse engineering and model siphoning efforts in real time, aiming to curb unauthorized proliferation.

• Governments are implementing export controls and regulations to restrict dual-use AI technologies, particularly autonomous weapons, while international efforts seek to establish binding treaties and norms to regulate autonomous systems and prevent destabilizing arms races.

• The U.S. government, citing national security concerns, has enacted measures such as President Trump’s order to cease federal use of certain models like those from Anthropic, which refuses to compromise on safety and ethical standards. Meanwhile, agencies are pivoting toward providers like OpenAI, which has secured Pentagon contracts, potentially reshaping supply chains and alliances.

Technological Enablers for Militarized AI

Advances in hardware and AI frameworks are facilitating rapid deployment of autonomous, weaponized AI systems:

• On-device model "printing" allows large models to be embedded directly onto specialized hardware, enabling autonomous units like drones and combat robots to operate independently in cyber-contested environments.

• GPS-denied navigation techniques are improving the reliability of autonomous hardware in battlefield conditions, critical for military operations under electronic warfare.

• Goal-directed, agentic AI systems such as Rover.ai are transforming decision-making in surveillance and targeting, further blurring civilian and military applications.

The Path Forward

The global landscape is now characterized by a fierce competition for AI hardware dominance, with private investments, national programs, and strategic collaborations shaping the future. While these developments promise technological advances, they also pose profound risks:

• Proliferation of militarized AI through illicit channels could escalate conflicts.
• Absence of robust international governance risks an uncontrolled arms race in autonomous weapon systems.
• The balance between responsible innovation and strategic advantage remains delicate.

Moving forward, the key challenge is establishing international norms, verification mechanisms, and cooperative frameworks that can prevent destabilization while fostering innovation. The decisions made today regarding AI chip development and infrastructure will determine whether humanity can harness AI for peace and progress or face an increasingly perilous future marked by autonomous warfare.