The AI Trail That Led Police to Two Missing Israelis — and What It Teaches About Privacy

The AI Trail That Led Police to Two Missing Israelis — and What It Teaches About Privacy In early August, the disappearance of Mali Yahalomi, 50, and her daughter Liel, 23, from the Modi'in area of Israel sparked an international search that stretched from Vienna to Buenos Aires. The pair were last in regular contact with family on a Friday evening while traveling through Europe, and their phones were later found abandoned near a Vienna train station. For a week, Israeli and Austrian authorities, alongside Interpol, worked to trace their movements. On a Friday night, Israel Police confirmed the two were found alive and safe in Argentina. But new reporting this week reveals a crucial detail in how they were located: their AI chatbot search history.

A Digital Trail Across Europe

According to N12 News, as reported by The Jerusalem Post, the mother and daughter were tracked down in Buenos Aires partly because of their interactions with an artificial intelligence chatbot. The pair had tried to cover their tracks by deleting the search history, but authorities recovered and traced enough information to pinpoint their location. The AI queries were remarkably specific. The two consulted the chatbot on how to buy a mobile device without showing identification, leading them to purchase a cell phone in the Czech Republic. From there, they traveled to Germany and eventually made their way to Argentina. The recovered history also included requests for advice on how to delete specific mobile apps to prevent location tracking, and which South American countries have no extradition agreement with Israel. These queries, pieced together, painted a clear picture of intent and route. The women were found on an intercity bus in Buenos Aires early Saturday morning, according to AJN, the Jewish news agency in Argentina. Israel Police representative in South America, Walter Alejandro Kogan, boarded the bus and located them, with a video of the encounter going viral. KAN News later reported the two had planned to change their identities and start new lives, with Mali telling Kogan she wanted a fresh start.

The Cyber Expert's Warning

For Osher Cohen, a cyber specialist and founder of Yo Secure, an Israeli company that handles digital identity crises, hacked accounts, online impersonation, and security incidents on digital platforms, the Yahalomi case is a stark illustration of a growing blind spot. Speaking to The Jerusalem Post, Cohen framed the incident not as a crime story, but as a privacy lesson for every user of AI tools. "Users need to gain a new habit: think before you type into an AI chatbot," he said. Cohen’s warning cuts to the heart of how Israelis—and indeed users worldwide—interact with the new generation of AI-powered search and chat tools. The Yahalomi case demonstrates that a conversation with a chatbot is not a private whisper. It is a data trail that can be recovered, analyzed, and used to reconstruct a person’s intentions, fears, and plans.

Chat vs. Search: A Richer Record

The distinction between a traditional search engine query and an AI conversation is central to Cohen’s concern. "A search can show what you wanted to know," he explained. "A conversation with AI can show what you were thinking, what you were worried about, what you were considering, and what you were trying to achieve." In the Yahalomi case, the difference was decisive. A search for "extradition agreements South America" might have been dismissed as a random query. But a chatbot conversation that included follow-up questions about deleting apps, buying phones without ID, and avoiding location tracking revealed a coherent plan. Cohen noted that the psychological comfort users feel with AI is part of the problem. "People sometimes treat AI chatbots as if they are speaking privately to themselves," he said. "In reality, an AI chatbot is a cloud service. It may be connected to an account, a device, an email address, a browser, an IP address, privacy settings, and data retention policies." The sense of intimacy with a conversational interface belies the fact that every prompt is processed through infrastructure that may log, store, or transmit data across multiple jurisdictions.

The Illusion of 'Just Searching'

The integration of AI into everyday search tools has blurred the line between looking something up and having a dialogue. "The biggest issue is psychological," Cohen told The Jerusalem Post. "When AI is built into search, people feel like they are 'just searching'. In practice, they may be creating a richer record than a regular search query." A standard search engine query is a string of keywords. An AI conversation is a narrative—complete with context, emotional tone, and a sequence of decisions. For law enforcement, that narrative is far more valuable than a list of links. The chatbot had effectively become a digital diary of their escape plan. Even after they deleted the history, authorities recovered and traced enough of the conversation to guide investigators to Buenos Aires.

Data Location and the Cloud Reality

Cohen also addressed a common misconception about digital privacy: that using a VPN or incognito mode makes a user invisible. He pointed out that large tech companies use global cloud infrastructure, and information may be processed or stored in different regions depending on the provider, product, account type, organizational settings, legal requirements, and internal architecture. VPNs help, he acknowledged, but many providers, networks, and apps may still have access to the data regardless of the user’s apparent location. The warning is especially relevant for Israelis: the country is home to major cyber companies and elite IDF intelligence units, and its citizens are generally more attuned to online threats than most. Yet the Yahalomi case shows that even tech-savvy individuals can underestimate the depth of the digital footprint they leave when interacting with AI. The pair clearly understood the need to delete apps and avoid location tracking—they just did not anticipate that the AI tool itself would become a key source of their location.

Behavioral Safety Over Technical Fixes

Cohen’s advice for users is grounded in a simple principle: "The safest approach is behavioral, not only technical." He urged users to avoid entering passwords, verification codes, or credit card details into AI chatbots. He also warned against uploading sensitive legal, medical, or business documents unless the user fully understands the risk. "Do not paste customer data into a personal AI account," he said, emphasizing that consumer AI tools are not designed to be secure repositories for confidential information. Practical steps, he suggested, include checking privacy settings, deleting history when it is no longer needed, and using temporary or incognito modes when appropriate. But the core message is about restraint. "An amazing tool, but it is not a safe deposit box," Cohen said. "The danger is not only what AI knows how to answer. The danger is what we choose to tell it."

A Lesson for Israeli Tech Culture

For Israel’s innovation economy, the Yahalomi case is a reminder that the same tools driving the country’s tech sector come with inherent risks. Israelis are early adopters of AI, using it for everything from coding to daily logistics. The country’s cybersecurity industry is world-renowned, but the consumer side of digital privacy often lags behind the enterprise level. Cohen’s comments suggest that the gap between professional security practices and personal habits is where vulnerabilities emerge. Cohen also offered guidance for businesses: companies should create an internal AI policy so employees know what they can paste into AI tools. Customer lists, contracts, legal documents, medical information, financial data, and internal strategy should not be casually uploaded into a personal chatbot account. In a country where startups routinely handle sensitive data, this advice carries particular weight. The same ease of use that makes AI so appealing in Tel Aviv’s offices can become a liability if employees treat corporate secrets as casual chat fodder.

The Recovered Trail and the Road Ahead

The Yahalomi case ended safely, with the two women found alive. But the method of their discovery has broader implications. Authorities recovered a digital trail that the women believed they had erased. The AI chatbot, which they used to plan their new lives, became the very tool that led police to them. For investigators, it was a breakthrough. For privacy advocates, it is a cautionary tale. As AI becomes more embedded in search, communication, and daily decision-making, the line between private thought and recorded data continues to blur. Cohen’s advice to "think before you type" is not just about avoiding incriminating queries. It is about understanding that every conversation with an AI is a conversation with a cloud service that may retain, process, and share that information. In a country like Israel, where digital security is a national priority, the lesson is particularly resonant: the most advanced technology is only as safe as the behavior of the person using it. The Yahalomi case will likely be studied in cybersecurity circles for years. It demonstrates that AI is not just a tool for answering questions—it is a tool for revealing the questioner. And for those who seek to disappear, the chatbot may be the last place they should turn.

This article was produced with AI-assisted research and editorial support. Sources: The Jerusalem Post, N12 News, KAN News, AJN, Israel Police.

By Hannah Berg, Staff Writer