America’s Endangered AI | Foreign Affairs

America’s Endangered AI | Foreign Affairs

America’s Endangered AI | Foreign Affairs

https://www.foreignaffairs.com/united-states/americas-endangered-ai

Publish Date: 2026-03-11 00:00:00

Source Domain: www.foreignaffairs.com

  • AI Models Used in Attacks: In 2025, U.S. companies like Anthropic and OpenAI reported large-scale cyber attacks using their AI models for espionage, illustrating the first documented case of AI-executed cyberattacks without substantial human intervention.
  • Increased Threats From Foreign Adversaries: Following U.S. strikes, Iran has intensified cyber operations leveraging AI to target Western infrastructure, prompting Palo Alto Networks to report over 60 Iranian-aligned groups using AI to exploit Western systems.
  • Vulnerability of U.S. Infrastructure: U.S. cyber defenses are inadequate, while AI rapidly integrates into American life and economic activities, increasing the potential damage from countries like China, Russia, and Iran.
  • Historical Cybersecurity Failures: For decades, Western companies have struggled against sophisticated espionage and intellectual property theft by Chinese actors, often missing their opportunities to protect critical technologies and innovations.
  • Strategic Importance of AI Infrastructure: AI infrastructure, featuring chips, data, software, and specialized hardware, attracts substantial interest from adversaries due to its economic and strategic value and the fragility of its security defenses.
  • Attacks on AI Systems: Hackers target AI systems aiming to steal or corrupt intellectual property, disrupt AI-powered services, and erode user trust by undermining system reliability during crises.
  • Immediate Measures Needed: U.S. policymakers must adopt strategies focused on improving defenses at chip, network/data, and human security levels to bolster national cyber-resilience and deter further cyber activities by foreign adversaries.
  • Holistic Approach for Strengthening Defenses: The U.S. should designate AI as critical infrastructure, combine lessons from successful foreign cyber strategies, and implement robust and collaborative measures involving industry and government to secure AI infrastructure.