The Information War Is Already Here — And Most Institutions Aren't Ready

Introduction

For most of my career in military public affairs, the information environment was shaped by human decisions — what to say, when to say it, and through which channels. The challenge was strategic: how do you communicate credibly in a complex, contested environment while maintaining institutional trust?

That challenge hasn't gone away. It has multiplied.

A growing body of peer-reviewed research and policy analysis confirms what many communicators have sensed for years: artificial intelligence has fundamentally changed the nature of the information environment. Not because AI created disinformation — humans have been doing that for as long as they have had language — but because AI has industrialized it. The scale, speed, and sophistication of AI-generated false content now exceeds the capacity of most institutions — governmental, journalistic, or military — to respond in real time.

Understanding this shift is no longer optional for people in strategic communication, public affairs, or any field where trust is the currency of effectiveness.

What the Research Actually Shows

A July 2025 peer-reviewed study published in PMC examined how AI-enabled tools — specifically generative models and engagement-optimization algorithms — are being used to produce and amplify disinformation at scale. The findings are sobering.

AI doesn't just create false content. It personalizes it. Engagement algorithms are optimized to find the emotional frequencies most likely to override critical thinking — outrage, fear, in-group loyalty — and they do so with increasing precision. Deepfake images, audio clips, and videos are now sophisticated enough that detection requires specialized tools most audiences don't have access to, and "pink slime" news networks — low-quality, AI-generated local news sites designed to appear legitimate — are seeding political ecosystems with fabricated content that is difficult to distinguish from journalism.

The research points to three primary mechanisms of harm. First, direct distortion: false narratives spread faster than corrections, and retraction rarely reaches the same audience as the original claim. Second, epistemic erosion: sustained exposure to disinformation doesn't just create false beliefs — it creates doubt about everything, making people less able to trust any source, including accurate ones. Third, polarization amplification: AI-driven content systems are optimized for engagement, and the most engaging content tends to be the most divisive.

Brookings Institution analysis adds a useful complication: AI is not inherently the enemy of democracy. It can also strengthen it — improving civic access, enabling translation, facilitating participation by historically excluded communities. The challenge is that the harmful applications are, at present, outrunning the beneficial ones.

Why Democratic Institutions Are Slow to Respond

This is not primarily a technology problem. It is a governance problem.

Democratic institutions are, by design, deliberate. They move through deliberation, legislation, public comment, and judicial review. These are features, not bugs — they protect civil liberties and prevent the arbitrary exercise of power. But they are poorly matched to the speed of AI development and deployment.

A regulation drafted in 2022 to address the disinformation landscape of 2020 will be obsolete before it passes committee. Platform accountability frameworks negotiated over years will be superseded by new architectures before the ink is dry. And the political will required to pass meaningful AI governance legislation tends to emerge only after significant harm — by which point the systems causing the harm are already embedded in the information environment.

The Inoculation Approach — and Its Limits

One of the more promising research findings concerns psychological inoculation — a communication strategy that pre-emptively exposes audiences to weakened forms of manipulation techniques so that they are better able to recognize and resist them when encountered in the wild.

Studies published in PMC and PNAS suggest that inoculation-style interventions can measurably improve discernment between authentic and false news, and can do so in ways that are scale-appropriate for social media deployment. The approach treats information resilience as a public health issue: just as vaccines train the immune system to recognize pathogens, inoculation techniques train the cognitive system to recognize manipulation.

The limits are real, however. Inoculation works better for technique-based manipulation — recognizing emotional appeals, false authority, conspiracy logic — than for specific false claims, which change faster than any inoculation protocol can track. And there is evidence that poorly designed media literacy interventions can backfire, creating generalized distrust rather than targeted discernment.

This means the human communicator, with contextual judgment and institutional credibility, remains irreplaceable — at least for now.

What This Means for Military Communicators

The U.S. military operates in and through the information environment. Its public affairs mission is not merely to push authorized messaging — it is to build and maintain the trust of the American public, international partners, and the service members themselves. That mission is increasingly complicated by an information environment saturated with AI-generated content designed to undermine exactly the kind of institutional trust military communication depends on.

Source credibility is the only asymmetric advantage. In a world where anyone can generate realistic-sounding content at scale, the institutions that invest in verified, attributable, consistently accurate communication will be the ones that maintain credibility when it matters. This is not a communication tactic — it is a strategic asset.

AI literacy is now a core competency. Every military communicator should have a working understanding of how generative AI works, what manipulation techniques AI enables, and what the research says about effective counter-messaging.

Closing Reflection

The information war is already here. The question is whether the institutions we trust to communicate with us honestly are equipped to fight it.

---

Sources

Romanishyn, Malytska, and Goncharuk. AI-Driven Disinformation: Policy Recommendations for Democratic Resilience. PMC. July 2025. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12351547/

Brookings Institution. How Do Artificial Intelligence and Disinformation Impact Elections? https://www.brookings.edu/articles/how-do-artificial-intelligence-and-disinformation-impact-elections/

Badawy et al. Computational Propaganda and AI Governance. PMC. 2025.

PNAS. Studies on psychological inoculation and media literacy. 2024-2025. https://www.pnas.org