Drooid Logo
Back to story perspectives

Full Breakdown

Global Arms Race: Conventional, Nuclear, and AI Dimensions

6/10/2026, 11:19:25 AM

Core Event

At the 23rd Shangri-La Dialogue, U.S. Secretary of War Pete Hegseth urged Asian nations to raise defence budgets to 3.5 % of GDP. A new report recorded global nuclear-arsenal spending at $119 billion in 2025, with the United States alone spending $69.2 billion. Banks and tech firms warned that AI tools are amplifying payment-fraud and spurring a cyber-security arms race, highlighted by Anthropic’s Project Glasswing and OpenAI’s confidential IPO filing.

Data & Statistics

Southeast Asian defence spending averages under 2 % of GDP; a 3.5 % target would force Indonesia, Malaysia, the Philippines and Thailand to more than double budgets. Nuclear outlays in 2025 total $119 billion, with the United States spending $69.2 billion and China $13.5 billion. About 12 000 warheads exist, roughly 9 700 ready for use. Visa flagged $1 billion in fraud attempts; banks spend $1-30 million on fraud tools. Anthropic’s Claude Mythos uncovered 10 000 software flaws. OpenAI is valued at $852 billion.

Official Statements & Responses

Hegseth warned that a “rules-based order is not worth the paper it is written on” without credible force, and Vietnamese President To Lam said defence industries must serve legitimate defence, not fuel an arms race. Malaysia’s defence minister observed the United States “plays a very strategic role… but every country faces different circumstances.” OpenAI’s confidential S-1 filing was described as giving “the option to go public sooner if that ends up being best.” The U.S. Payments Forum issued guidance for banks on AI-driven fraud.

Criticism, Opposition & Gaps

Southeast Asian leaders argued that security rests on stability and relationships, not spending spikes. Ican’s Susi Snyder called the nuclear-spending surge “deeply alarming” and “terrifying.” Bank executives warned that “yesterday’s tools are a lot more ineffective” against AI-enhanced fraud, while BeyondTrust cautioned that proliferating powerful AI models could enable large-scale cyberattacks if misused.

Verbatim Quotes

  • “those who long for peace must prepare for war.” — Pete Hegseth, U.S. Secretary of War
  • “We respect the call by the United States… they play a very strategic role to create a balance in the Indo-Pacific. But they must also understand every country faces different circumstances… If America asks NATO and Europe to increase their spending, we can understand. But for a country like Malaysia, a developing country, we have other sectors we need to develop.” — Malaysia’s defence minister
  • “To be perfectly honest, I’m terrified,” — Susi Snyder, Ican director of programmes
  • “AI made fraud economics work a lot more effectively at scale,” — Saurabh Bajaj, chief product officer, Oscilar

Why It Matters / Impact

Higher conventional spending could strain health, infrastructure and climate-resilience budgets in developing economies. Record nuclear outlays raise the danger that AI-enhanced decision-making shortcuts may trigger accidental use. AI-driven fraud threatens financial stability, while the cyber-security arms race risks a backlog of unpatched vulnerabilities that could compromise critical services worldwide.

What’s Next

Southeast Asian governments will review defence-budget proposals before the next fiscal cycle. The United States and China have outlined multi-decade nuclear-modernisation plans. Regulators are drafting AI-risk frameworks, and Project Glasswing aims to add more than 150 partners by year-end.