The Next Information War Will Be Fought Inside AI

The battlefield has moved from attention to memory.

For the past decade, the battle over public opinion has taken place on social media. Governments, activists, journalists, and advocacy groups have fought over hashtags, viral videos, trending topics, and online narratives. Whoever shaped the conversation on X, Facebook, Instagram, TikTok, or YouTube seemed to shape reality itself. That era is passing.

The next information war will be fought inside artificial intelligence, and instead of having an influence that last as long as the news cycle, it will change the way we define truth for generations.

The coming information war will be a battle to influence what AI systems learn, remember, summarize, and present as knowledge and truth.

Not long ago, those seeking to shape public opinion focused on journalists. Get the headline right, influence the editor, frame the news segment, win the television interview, dominate the social media cycle. That strategy still matters. But artificial intelligence introduces a new and deeper target because AI does not merely report the news, it digests the archive.

Large language models absorb not only data, but patterns of emphasis, vocabulary, causality, and moral framing. What appears repeatedly across the digital record increasingly shapes the way AI explains the world. This transitions the significance of information from what people read today to what the machines learn for tomorrow.

This changes the nature of influence. A viral post may last a day. A misleading headline may shape opinion for a week. But narratives embedded across digital archives will influence AI systems for generations, resurfacing through search engines, AI summaries, classroom materials, and automated research tools. The battlefield has shifted from controlling today’s headlines to shaping tomorrow’s machine memory.

Digital archives therefore take on an entirely new significance. Newspapers, think tanks, universities, NGOs, governments, and international organizations collectively create the historical record from which future AI systems learn. If those archives consistently describe one side of a conflict in legal terminology while portraying the other in emotional language, AI learns that distinction. If certain atrocities receive exhaustive documentation while others are treated as background noise, AI learns that imbalance. If historical context is repeatedly omitted, the omission itself becomes part of the machine’s understanding of history.

Among the most strategically important battlegrounds is Wikipedia. For millions of people, Wikipedia is the first stop for basic information. For AI systems, it is a very valuable, structured, cross-referenced, constantly updated knowledge backbone. Whoever influences the framing of Wikipedia entries today influences the informational scaffolding on which AI systems rely. The forces seeking to undermine Israel, and through it critical elements of Western civilization understand this and have made Wikipedia a central focus of their disinformation assault.

This is why edit wars matter. A disputed phrase, a missing source, a loaded label, or an imbalanced chronology may seem trivial to some, but over time these details shape the architecture of public knowledge. AI systems trained on or connected to those sources will reproduce these distortions, fabrications, and half-truths at global scale.

Public datasets are another critical front. AI developers rely on large collections of text to train, test, and evaluate their models. These datasets often include material scraped from the open web, curated from academic sources, or drawn from public repositories. Once information enters such datasets it contributes to the statistical patterns through which AI models understand truth.

This is where advocacy enters an entirely new phase. In the past, organizations focused primarily on influencing today’s conversation. Increasingly, they must also think about influencing the long-term body of knowledge from which tomorrow’s AI systems learn.

Online news is central front in this battle. Every article published today has the potential to become part of tomorrow’s informational environment. Reports that omit essential context, headlines that oversimplify complex events, or stories that privilege emotional impact over fact do not necessarily disappear when the news cycle ends. They remain searchable, quotable, archivable, and potentially incorporated into AI systems.

This permanence should make journalists more cautious, not less. But the incentives shaping journalism have changed. Reporters no longer write solely for editors. Today, they write for algorithms and audiences most likely to give a thumbs up and a share. Social media rewards outrage, certainty, speed, and engagement far more than nuance, complexity, or patience. These incentives have influenced coverage. The challenge is that what once produced only temporary distortions now risks creating permanent ones.

For Israel, these developments carry especially serious implications. The Jewish state has long been the subject of intense media scrutiny, disproportionate attention, and recurring double standards. In the age of AI, these distortions are absorbed into the AI systems that increasingly mediate how students, policymakers, journalists, and ordinary citizens understand the world. Israel is therefore becoming one of the earliest test cases for understanding how AI inherits human bias. What happens here today may foreshadow how AI will treat many other controversial subjects tomorrow.

Ask an AI assistant about Zionism, Gaza, settlements, terrorism, international law, or the history of the Arab-Israel conflict, and it will synthesize its response from the information available to it. If that information reflects persistent imbalance, the synthesis will reflect persistent imbalance as well. The danger is that AI will become a highly polished echo of accumulated imbalance.

This is why the old model of media criticism is no longer enough. Correcting a headline, demanding an editor’s note, and challenging a misleading caption still matter. But protecting the future increasingly requires improving the digital record itself.

This means strengthening reliable public archives, contributing to high-quality datasets, engaging seriously with Wikipedia and other knowledge platforms, preserving historical context, documenting events accurately, and  ensuring that context, chronology, law, and moral clarity are preserved in the searchable digital world for AI systems to encounter.

Today, many organizations concerned with Israel, antisemitism, democratic values, and the integrity of public discourse remain organized to fight yesterday’s information war. They devote enormous resources to rebutting headlines, responding to social media trends, and correcting immediate misinformation. Those efforts remain important, but they are no longer sufficient. In order to be fully prepared for the next information war must now proactively confront the battle for the permanent knowledge infrastructure from which AI systems learn.

This is the new frontier of information warfare. It is less visible than a trending hashtag, less dramatic than a viral video, and less glamorous than a television debate. But it may prove far more enduring.

The next generation will not ask what a newspaper said about Israel. It will ask an AI system to explain Israel. The answer will depend on what the machine has learned from our archives, our encyclopedias, our datasets, and our journalism. Accuracy is no longer only about correcting the present, but about protecting the future.

The next information war will be fought inside AI because AI is becoming the world’s emerging memory. Whoever shapes that memory will shape how future generations understand democracy, history, and ultimately, truth itself.