Japanese Prime Minister Sanae Takaichi and U.S. President Donald Trump held talks in New York on September 22, just days before a high-profile summit between Trump and Chinese leader Xi Jinping in Washington on September 24. The meetings underscored the evolving complexities in the strategic relationship among the United States, China, and their key allies, including Japan.

Japan entered the discussions seeking a robust alignment with the United States on security concerns in the East and South China Seas, as well as on maintaining stability in the Taiwan Strait. Tokyo emphasized the importance of a firm stance against any unilateral attempts by China to alter the status quo in these sensitive regions. However, Washington appeared to prioritize an emerging area of concern over these territorial disputes: the risks posed by artificial intelligence (AI) to military security and global stability.

While trade issues such as critical minerals—including rare earth elements—and tariffs were also on the agenda, Japan expressed reservations that the Taiwan issue and U.S. arms sales to Taiwan might not receive sufficient attention during U.S.-China negotiations. The bilateral talks between Takaichi and Trump addressed a range of topics including economic security, semiconductor supply chains, and AI, reaffirming the strength of the U.S.-Japan alliance amid rising regional tensions.

At the U.S.-China summit, there was limited movement toward resolving disagreements over trade, mineral resources, or Taiwan. Instead, AI emerged as a focal point with potentially far-reaching implications beyond immediate technological competition. Experts suggest that AI could evolve into a form of strategic deterrence akin to nuclear weapons, possibly transforming global security dynamics within the next two decades.

Drawing parallels to Cold War nuclear deterrence, some analysts warn that AI capabilities might eventually enable the United States and China to disrupt or neutralize each other’s strategic systems. This prospect raises the possibility of an AI-based version of “mutually assured destruction” (MAD), where the recognition of catastrophic retaliation risks might deter direct conflict. However, neither government is currently willing to impose meaningful limits on AI development due to the strategic advantages at stake.

The emerging AI competition is poised to coincide with a shift in the nuclear balance of power. China’s nuclear stockpile is projected to expand significantly, potentially surpassing 1,500 warheads, a threshold that could establish it as a third nuclear superpower alongside the United States and Russia. This tripolar nuclear framework could complicate traditional deterrence models originally formulated during the bilateral U.S.-Soviet rivalry, making strategic calculations more unpredictable.

A similar complexity may arise in AI deterrence if Russia becomes a significant actor in this domain alongside the United States and China. The potential for a multipolar AI balance raises questions about the stability of established security frameworks.

Observers note that the initial AI discussions at the recent summit may signal early shifts in the broader great-power balance. For Japan and other Asian countries, recognizing the strategic implications of AI competition alongside more conventional issues like Taiwan will be critical in navigating the region’s evolving security landscape in the coming decades.