Disinformation Amplifies Global Existential Threats, Study Warns

In the age-old battle against truth, AI emerges as a powerful new weapon in deception’s arsenal, experts warn. What began with wooden horses has evolved into digital Trojans of unprecedented sophistication that threaten global stability and nuclear security.

Deception as a military and political tactic dates back millennia. In the 12th or 13th century BCE, Greek forces employed what became history’s most famous ruse: the Trojan horse. By constructing a massive wooden horse containing hidden soldiers and abandoning it outside Troy’s gates, they tricked their enemies into bringing their own destruction within their walls. The stratagem was so effective that “Trojan horse” remains a common phrase today, particularly in computing security, where it describes malicious software disguised as legitimate programs.

While deception itself isn’t new, the sophistication, scale, and potential consequences of modern disinformation campaigns have reached unprecedented levels. Artificial intelligence has dramatically accelerated the creation of convincing deepfake videos, synthetic content, and fabricated accounts, overwhelming traditional fact-checking mechanisms.

“AI brings a significant possibility of elevating nuclear escalation risks by amplifying disinformation, overloading analysts, compressing decision timelines, and exploiting cognitive and institutional vulnerabilities in sociotechnical systems for nuclear command and control,” writes security analyst Herb Lin in a recent analysis for the Bulletin of the Atomic Scientists.

Lin presents three scenarios illustrating how AI could function as a “threat multiplier” in nuclear tensions—by distorting perceptions, contaminating intelligence, and disrupting the delicate signaling mechanisms nuclear powers use to communicate intentions. His detailed analysis reveals how rapidly such technology could push international crises beyond control points.

These scenarios aren’t merely theoretical. In October 2022, according to researcher Polina Sinovets, Russian President Vladimir Putin deployed deepfakes and disinformation to falsely claim Ukraine was preparing to detonate a “dirty bomb.” This fabricated narrative appeared designed to provide preemptive justification for Russia’s potential use of tactical nuclear weapons during a critical phase of the Ukraine conflict when Russian forces were struggling to withdraw 20,000-30,000 troops from southern Kherson across the Dnipro River.

The impact of sophisticated disinformation extends far beyond military domains. Society-wide erosion of trust in evidence and expertise threatens effective responses to challenges ranging from public health to climate change. Without shared reality, conspiracy theories flourish despite contradicting established facts.

The consequences have already manifested in governance. Lisa Fazio notes that “The US Department of Health and Human Services is now run by conspiracy theorists who believe that the American public health system is hiding key data on vaccine safety and who spend their days spreading health misinformation.”

Despite these challenges, experts suggest that countering disinformation, while complex, remains possible through multi-faceted approaches addressing supply, demand, distribution, and consumption of false information. Improving information environments can still help people make better-informed decisions, even as social media platforms retreat from fact-checking responsibilities.

Joseph Uscinski, political science professor at the University of Miami and author of several books on conspiracy theories, advocates for direct confrontation of misinformation, even when it comes from friends and family. “Being tolerant and compassionate [about disinformation-riddled conspiracy thinking] isn’t the same as pretending that their behavior isn’t their behavior,” he states. “I have compassion for them, but I hold them responsible for their beliefs and behaviors.”

As AI capabilities continue advancing, the battle between truth and deception enters a new phase where the stakes include not just public discourse but potentially nuclear stability. The modern Trojan horse doesn’t arrive as wood and soldiers but as pixels and algorithms—yet its danger to protected institutions may prove even greater than its ancient predecessor.