An Iranian group, Storm-2035, exploited ChatGPT to generate content for influencing the US presidential election, according to The Guardian. The exploitation of ChatGPT by Storm-2035 blurs the line between open-source AI and state-sponsored cyber warfare, enabling rapid, scalable disinformation campaigns.
The West develops powerful, accessible AI models, yet adversaries like Iran weaponize these same tools to undermine Western interests. Iran's military already harnesses ChatGPT for its cyber operations, reports the Financial Times. Iran's military harnessing ChatGPT for its cyber operations creates a critical security liability, turning a tool of progress into a weapon against its creators.
The rapid adoption of AI by state actors like Iran for both influence and offensive cyber operations will make the global cyber landscape increasingly complex and asymmetric. Urgent policy and defense innovations are demanded to mitigate these emerging threats.
How Iran's AI-Powered Influence Operations Work
- Iran's operation leveraged ChatGPT to generate content on US election candidates, the Gaza conflict, and Israel's Olympic presence, according to The Guardian. The versatility of Iran's operation, spanning long-form articles and short social media comments, demonstrates AI's capacity to automate and diversify influence operations across multiple platforms. Such automation makes detection and attribution significantly more challenging, amplifying the threat of foreign interference.
Beyond Influence: AI for Direct Cyber Attacks
Iran's military is now using Western AI models to develop malware and launch cyber attacks, reports the Financial Times. Iran's military using Western AI models to develop malware and launch cyber attacks represents a dangerous escalation from information warfare to direct, potentially destructive cyber attacks. The open accessibility of powerful, general-purpose AI models drastically lowers the barrier to entry for advanced cyber warfare, allowing state actors like Iran to rapidly scale sophisticated operations without developing their own foundational AI technologies. The open accessibility of powerful, general-purpose AI models fundamentally alters the cost-benefit analysis for state-sponsored cyber aggression.
The Broader AI Arms Race and Geopolitical Tensions
Anthropic's AI tools are integrated into US military classified networks, according to Vox, demonstrating America's commitment to leveraging advanced AI for defense. Yet, concurrently, Western AI models are being used by Iran's military to develop malware and launch cyber attacks, as reported by the Financial Times. The paradox of Western AI models being used by both US military for defense and Iran's military for cyber attacks fuels an arms race where the West inadvertently distributes the very tools of conflict. Both sides are rapidly integrating advanced AI into military strategies, signaling a new, technologically driven phase of global confrontation, where the lines between innovation and threat are increasingly blurred.
The Escalating Risks of Accessible AI
The open availability of powerful Western AI models like ChatGPT inadvertently fuels an asymmetric cyber arms race. Adversaries can rapidly enhance capabilities without significant R&D investment, leveraging tools developed for progress against their creators. The ease of weaponization, allowing adversaries to rapidly enhance capabilities without significant R&D investment, demands urgent international frameworks and robust defensive measures to prevent widespread misuse and unintended escalation.
If stringent safeguards are not implemented by major AI developers, it appears likely that governments will impose regulations to prevent the exploitation of these models for cyber warfare by Q4 2026.
