Israel's $610 billion economy is functionally invisible inside AI search engines like Perplexity and ChatGPT, even as these platforms become the primary way younger generations discover information about global conflicts and policy. This visibility gap represents a new kind of diplomatic vulnerability that traditional public relations strategies cannot address, according to analysis of how modern answer engines retrieve and cite sources.

The problem emerged as Israeli officials began acknowledging a bipartisan erosion of support among American voters. Gilad Erdan, former Israeli ambassador to the United Nations, told The Jerusalem Post this week that Israel is "losing the young generation of the Republican Party," a leading indicator that the Israeli government has not publicly acknowledged. That shift signals a fundamental change in how information shapes foreign policy perception.

How Do AI Search Engines Actually Decide What to Show You?

When you ask Perplexity or ChatGPT a question about geopolitics, the answer you receive travels through a multi-stage pipeline that most users never see. The engine first decides whether to search the web at all, then reformulates your question into multiple hidden searches, retrieves candidate sources, reranks them based on relevance and other signals, and finally selects which sources to cite in the answer.

The critical insight: being indexed by Google is not the same as being retrieved by an AI engine; being retrieved is not the same as being placed into the model's context window; and being in context is not the same as actually shaping the final answer. This multi-stage filtering means that visibility in traditional search does not guarantee visibility in generative AI search.

A 2026 study of Google AI Overviews found that nearly 30% of cited domains did not appear in the co-displayed first-page search results, indicating that the source-selection path for generative answers operates on a fundamentally different logic than conventional web ranking. For Israeli institutions and companies, this means they can rank well in traditional Google Search and still remain invisible to the AI engines that now answer questions for millions of users.

Why Is Israel's Economy Invisible to AI Engines?

The Olam Index, described as the first AI-visibility audit of the Israeli economy, found that seven of the ten most AI-visible Israeli companies are cybersecurity firms, and the rest of the economy is functionally invisible to generative search. Israeli tech companies maintain a $30 billion footprint in the United States. Israeli defense-tech firms are signing record commercial leases. Yet when an AI engine searches for information about Israeli innovation, technology, or economic capacity, it cannot reliably find or cite these institutions.

This invisibility stems from how AI engines retrieve sources. The retrieval process relies heavily on English-language content that has already been indexed and deemed relevant by the underlying search systems. Israeli newsrooms, government institutions, and tech companies produce content primarily in Hebrew or in English-language outlets with limited reach. When that content does get translated or recycled through the AI pipeline, it often loses attribution and context.

The result is that when a college student asks ChatGPT or Perplexity about Israeli innovation, defense capabilities, or economic strength, the answer is shaped by sources that are overwhelmingly Western media outlets, not Israeli institutions themselves. The narrative is filtered through a lens that Israeli officials have no direct control over.

What Would a Real AI-Era Communications Strategy Require?

Erdan's proposal for an "army of hasbara" (the Hebrew term for public diplomacy) is framed in traditional terms: media, campuses, diplomats, and budgets. But addressing the AI visibility gap would require capabilities that most governments have not yet built:

• Generative Engine Optimization (GEO): Building the retrieval anchors, entity signals, and citation infrastructure that shapes what AI engines say about Israel across every topic vertical, from defense technology to economic innovation to cultural institutions.
• Structural investment in English-language source material: Creating original reporting, research, and analysis that AI engines retrieve and cite, not promotional pamphlets or op-eds, but institutional content built specifically for the AI layer.
• Permanent digital infrastructure: Operating year-round campus and digital presence that builds citation share before the next news cycle, rather than reactive campaigns spun up after a crisis.
• Private sector integration: Coordinating with Israel's leading PR firms, tech companies, and academic institutions to function as a national communications ecosystem, not a government-only function.
• Institutional continuity: Establishing a single command authority with a multi-year mandate and real budget, protected from the coalition politics that have repeatedly defunded and repurposed Israeli communications ministries.

The challenge is structural. Israeli governments have historically treated public diplomacy as a side function, distributed across the Foreign Ministry, the Government Press Office, and rotating political appointees with no institutional continuity. The Strategic Affairs Ministry was created, defunded, restored, repurposed, and effectively shelved within a decade. No communications operation can function inside that cycle, let alone one designed to compete for visibility inside AI engines that operate on timescales measured in milliseconds.

Why Traditional Hasbara Strategies No Longer Work Alone

Erdan's diagnosis of the problem is accurate but a decade old. Israel's public diplomacy budget has been fragmented across five government ministries with no unified doctrine since at least 2016. The Foreign Ministry and Tourism Ministry have recently hired American PR firms, including a Florida ad shop for paid media and Mercury Public Affairs for Washington advocacy, according to FARA filings from July 2026. But booking agencies is not building infrastructure. Two retainers do not constitute a strategy.

The deeper issue is that the battlefield has moved. The narrative war is no longer primarily fought on CNN, campus newspapers, or the UN General Assembly. It is fought in the answer box inside ChatGPT, in the citations that Perplexity selects, in the sources that Claude retrieves when a user asks about Middle Eastern geopolitics. An "army of hasbara" without an AI communications strategy is a 2015 solution applied to a 2026 problem.

When search activation occurs, the AI engine's retrieval system pulls from a candidate pool shaped by indexing, crawling, and ranking systems that favor English-language, Western-published content. Israeli institutions that do not actively build presence in that layer remain invisible, regardless of their actual economic or strategic importance. The result is that younger generations asking AI engines about Israel receive answers shaped by sources that Israel's government did not select and cannot easily influence.

Whether any Israeli government will actually build the institutional infrastructure required to compete in the AI era remains an open question. The diagnosis is clear. The prescription is waiting for a government willing to fund it, staff it, protect it from coalition politics, and modernize it for the answer-engine era.