Back to articles

Israeli business group gives Meta 30 days to act on fake Zara, Castro and Aldo store ads

What happened

Israel's Federation of Chambers of Commerce has asked Meta to move faster against sponsored Facebook and Instagram ads that impersonate well-known retailers and send shoppers to fake stores. Federation president Shachar Turgeman sent the letter to Adi Soffer Teeni, chief executive of Meta Israel, and ynet reported it on Monday, November 16, 2026. The letter describes commercial identity theft and consumer fraud carried out through paid advertising, and asks Meta to reply within 30 days with the practical steps it plans to take (ynet).

The reports are in Hebrew. Quotes and summaries in this article are our own translations and paraphrases.

Fake stores under familiar names

According to ynet, sites impersonating Zara, Delta, Castro, Aldo, Tamnoon, Yad2 and Shar HaMashkaot recently appeared and promoted deep discounts that did not exist. Shoppers who tried to buy were asked for card details, which the operators stole. The report adds that Shufersal, Osher Ad, Bank Hapoalim, Israel Post, the Israel Electric Corporation, Adidas and Rami Levy have also been impersonated in similar schemes (ynet). The business site ice described the retail cases the same way: links to fake shopping sites spread on social networks, unusual discounts and attempts to collect card details (ice).

Neither report says how many shoppers were affected, how much money they lost, or whether the cases share operators. The list itself is telling. It mixes fashion chains, a supermarket chain, a classifieds site, a bank, the postal service and the electricity company. What these names share is trust. A familiar logo and an offer that seems almost plausible are enough to win a quick click.

Turgeman gave ynet one example. A few weeks earlier, he said, someone copied Aldo's entire website and offered customers 80% off. In his account it took Facebook three weeks to respond, and even then the response came through personal connections. In the meantime the brand's good name suffered and customers were angry at the business, which had no good way to answer them (ynet). Neither Aldo nor Meta has published its own account of that case, so the timeline rests on Turgeman's description.

What the federation asked for

The letter makes two requests. The first is better detection and prevention of impersonating ads, with fewer repeat appearances. The second is a clear, accessible and fast channel for businesses that discover someone operating in their name. The federation says that channel should have dedicated resources, allow direct reporting of impersonation, deliver an initial review within a few hours, and allow immediate action to stop suspicious ads when there are real signs of fraud, without waiting for a long inquiry to finish (ynet).

Turgeman compared the model he wants to the way credit card companies handle fraud: act quickly once fraud is flagged, then look into it. He also pointed to Meta's ad targeting. A company that can match an ad to the right user, he argued, should be able to stop more fraudulent ads before they run, and should catch the rest quickly through a dedicated complaints process (ynet).

The letter is careful about scope. It says the federation is not asking Facebook to police the internet, does not see Meta as a partner in the fraud, and does not favor sweeping liability for the acts of third parties. It also warns that if impersonation keeps spreading without an effective answer, calls for legislation and regulation will grow louder. Turgeman told ynet the federation would go to lawmakers if it had to, although it would prefer that Meta regulate itself.

One passage captures the business side of the problem. Behind every impersonating ad, the letter says, there may be two victims: the shopper who believed they were buying from a trusted store, and the real business left to handle complaints about a sale it never made. The ad is fake and the sale is fictitious, but the damage to reputation is real.

Meta's response

Facebook Israel told ynet it takes scams very seriously and continues to invest in people, technology and partnerships to find and act against them. It said it removed more than 159 million scam ads last year, that fraudsters keep changing their methods and work across many apps and platforms at once, and that it would be glad to work with the federation and with Israeli businesses. It called the claim that it accepts this activity, or avoids acting against it, wrong (ynet).

The 159 million figure matches what Meta announced in March 2026. At the same time, Meta said it would expand advertiser verification so that verified advertisers account for 90% of its ad revenue by the end of 2026, up from about 70%, and that it had built AI systems to flag more ads impersonating brands and public figures (WIRED). Since August 2025, Meta has also said that businesses using its Brand Rights Protection tool can report suspected scam ads that use their brand name without permission, even when the ad does not directly copy their intellectual property (Press Gazette).

Bank Leumi's lawsuit in the background

The letter comes three months after Bank Leumi sued Meta. On August 17, 2026, the bank filed a NIS 26.6 million claim in the Tel Aviv District Court, alleging that Meta runs sponsored ads that impersonate the bank and offer fake benefits to collect login details and verification codes. The bank said it identified 243 fraudulent ads in July, more than in the whole first half of 2026. It asked the court to require identity checks for every advertiser, warnings on ads from unverified accounts, and access to ads that use its name before they run (Calcalist). Meta disputes the claims, and the case is ongoing (ISOC-IL). The federation's letter cites the suit as an example of how widespread the problem is, while saying it takes no position on the legal arguments.

On Monday evening, after ynet's report, Leumi chief executive Hanan Friedman posted on X that the list of Facebook's victims is growing and that the platform keeps profiting from fraud, according to ice. ice also reports that Friedman criticized Meta a month earlier over ads impersonating journalist Shaul Amsterdamski and financial figures (ice).

What remains unknown

Several basic facts are not public. The reports do not name the domains, ad accounts or pages behind the fake stores, or say whether one group ran more than one of them. They do not say how many ads ran, how long each stayed live, or how much shoppers lost. The three-week response time in the Aldo case is one person's account. Meta's 30-day window runs to mid-December, and neither side has published the full letter or a formal reply.

Why this matters beyond Israel

The pattern in the letter is familiar far beyond Israel. A paid ad borrows a brand's name and images, links to a copy of its store, and collects payment details from people who trust the name. The real business absorbs the complaints, the angry comments and the hours spent reporting. The federation's requests double as a checklist any brand can use on itself: how quickly it can find an ad that uses its name, how fast it can get that ad and the store behind it removed, and what it can tell customers while that happens.

Where Outtake could have helped

This section is hypothetical analysis, kept separate from the reported facts above. Nothing in the reports says whether any of the named brands use an outside monitoring service, and nothing here suggests that any of them use or lack Outtake. It maps the parts of this pattern that are visible from outside to capabilities Outtake describes publicly.

Watching paid ads for the brand's look, not only its name

Fake retail ads often use the brand's logo, product photos and colors, while the advertiser name and the link look generic. Outtake Brand Protection monitors paid advertising networks alongside domains, social platforms and app stores. It is designed to spot fake ads by visual signal rather than keyword match, across text, images and video. For a retailer, that could mean a sponsored post using its storefront images or a current campaign photo gets flagged even when the ad never spells out the brand name. Our guide to cloned and malicious ads covers the signs in more detail.

Following the ad to the store behind it

The ad is the entry point. The fake store is where card details are taken. Removing only the ad leaves the store live for anyone who saved the link or reaches it through another ad. Monitoring that covers both ads and domains could connect a flagged ad to the lookalike store it promotes and capture evidence from both while they are live, including the copied pages and the checkout form that asks for card details.

Linking one clone to the next

Turgeman described a whole site being copied. Copied stores often reuse something: a template, hosting, a registration pattern, analytics or tracking IDs, or the social accounts that promote them. Outtake says that from a single flagged domain its agents trace shared hosting, registration patterns, analytics IDs and linked social accounts (Brand Protection), and its Recon Agent is built to expand one signal across the connected network. If several of the fake stores named here came from the same kit, which the reports do not establish, that kind of tracing could surface related sites together instead of one report at a time.

Building a report a platform can act on

The federation wants Meta to review credible reports within hours. Whatever Meta decides, a reviewer can usually act faster on a complete report: which ads, which pages and accounts, which landing domain, what was copied, and how it differs from the real business. Outtake runs remediation through ad networks, social platforms, registrars and hosts (Brand Protection). A report that bundles the ad, the landing page and the registration details may leave less for the platform to investigate. It cannot set the platform's queue or its pace.

Watching for the same fake to return

The federation's first request was not only faster removal but fewer repeat appearances. A removed ad can be replaced by a new one from a different account, and a suspended store can reopen on a fresh domain with the same pages. Outtake's Recon Agent page says it picks up where takedowns end, looking for infrastructure that survived and new assets that appear after removal. For a brand, that could mean checking whether the template, images or tracking IDs from a removed store show up again under another name, and reporting the new copy as part of the same case rather than starting over. Whether the Israeli fakes reappeared after removal is not reported.

Getting the signal to the people who answer customers

Turgeman's complaint was also about the gap after discovery, when customers are angry at the brand while it waits. Outtake sends findings into SIEM, SOAR and ticketing tools so that marketing, communications, legal and security teams work from the same record (Brand Protection). A support team that knows which fake domains are live could answer complaints with specifics and post a warning that names them in safe, non-clickable form, rather than a general caution. Our guide to brand impersonation on social media and marketplaces covers how teams split that work.

Trace a fake retail ad from the feed to the platform's response

Choose a path, then step through its stages. Each stage shows what the reports say, where outside monitoring could have intervened (hypothetical analysis, not a reported event) and the limit of that intervention. Reported details come from ynet, ice, Calcalist and Meta's public statements, translated from Hebrew where needed.

What the shopper sees

Sponsored ad

A sponsored post shows a famous brand and a deep discount

Reported evidence

Sites impersonating Zara, Delta, Castro, Aldo, Tamnoon, Yad2 and Shar HaMashkaot promoted deep discounts that did not exist, and the federation says they reached shoppers through sponsored Facebook and Instagram ads (ynet).

Possible external intervention

Hypothetical: monitoring of paid ads that matches logos, product photos and layout could flag the post even when the ad text avoids the brand name (Brand Protection).

Control boundary

Ads targeted at a narrow audience may not be visible to outside monitoring before shoppers see them.

Cloned store

The link opens a copy of the store

Reported evidence

Turgeman told ynet that someone copied Aldo's entire website and offered customers 80% off (ynet). The reports do not name the domains used.

Possible external intervention

Hypothetical: linking the flagged ad to its landing domain could classify the store as a clone by how it looks and capture evidence of the copied pages while they are live.

Control boundary

Because the domains are not public, it is unknown how early they could have been found, or whether they showed the same page to every visitor.

Checkout

The shopper enters card details

Reported evidence

Both reports say the fake stores were built to steal card details from shoppers who tried to buy (ynet, ice).

Possible external intervention

Hypothetical: the main outside lever is removing the ad and the store so fewer people reach the checkout page at all.

Control boundary

Details typed into the form reach the operators at once. Refunds and blocked cards are handled by banks and card issuers, not by takedowns.

Aftermath

Angry customers turn to the real brand

Reported evidence

Turgeman said the brand's good name suffered and customers were angry at the business, which had no way to answer them. The letter describes two victims: the shopper and the real business (ynet).

Possible external intervention

Hypothetical: a shared record of which fakes are live could help support and communications teams answer complaints with specifics and warn customers early.

Control boundary

No monitoring service can undo a shopper's loss or reach shoppers who never contact the brand.

What the brand can do

Discovery

The brand learns a fake is running

Reported evidence

The reports do not say how the retailers found the fake stores. In its lawsuit, Bank Leumi said it identified 243 fraudulent ads impersonating it in July 2026, more than in the whole first half of the year (Calcalist).

Possible external intervention

Hypothetical: continuous monitoring of ad networks and new domains could move discovery earlier, so it comes from monitoring rather than from customer complaints.

Control boundary

There is no public data on how quickly these retailers found the fakes, so any time saved is speculative.

Report

The brand reports the ad and the page to Meta

Reported evidence

Meta has said businesses using its Brand Rights Protection tool can report suspected scam ads that misuse their brand name (Press Gazette). The federation is asking for a dedicated, direct channel for impersonation reports (ynet).

Possible external intervention

Hypothetical: a report that bundles the ad, the page, the landing domain and its registration details may give a reviewer less to investigate.

Control boundary

Whether and when the ad comes down is Meta's decision.

Wait

The report waits for review

Reported evidence

Turgeman said Meta took three weeks to respond in the Aldo case, and then only through connections. The federation asks for initial review within a few hours (ynet).

Possible external intervention

Hypothetical: while the ad waits, the brand can pursue the landing domain through its registrar and host, which Outtake lists among its remediation routes (Brand Protection).

Control boundary

Registrars and hosts apply their own policies and timelines, and a removed store can reappear on a new domain.

Escalation

Escalation moves to letters, lawsuits and lawmakers

Reported evidence

The federation set a 30-day deadline and raised the prospect of legislation. Bank Leumi filed a NIS 26.6 million suit against Meta in August, and its chief executive criticized Meta again on November 16 (ynet, Calcalist, ice).

Possible external intervention

Hypothetical: a dated record of each ad, report and response could support escalation with a platform, an industry body or counsel.

Control boundary

Evidence records do not settle legal questions. The Leumi case is ongoing and Meta disputes the claims.

What outside monitoring cannot reach

Monitoring sees only what is public, and several parts of this pattern are not. Narrowly targeted ads may reach a small audience and never be seen by an outside crawler before shoppers see them. Some fraudulent ads change their landing page after review or show different content to different visitors, so evidence has to be captured while it is live. Card details typed into a fake store reach the operators immediately. Getting that money back runs through banks and card issuers, not takedowns.

The decision to remove an ad, a page or an account belongs to Meta, and the decision to suspend a domain or a hosted store belongs to the registrar or host. Their response times are their own, which is the federation's core complaint. Takedowns also do not identify or stop the people behind a scheme. That takes law enforcement and, as the Israeli debate shows, sometimes courts and lawmakers. A brand can shorten the time it takes to find and report a fake. It cannot make a platform answer in hours.

Frequently asked questions

Were Zara, Castro or Aldo hacked?

No report says so. The reports describe impersonation: separate fake sites that used the brands' names and promoted fake discounts through sponsored ads (ynet, ice).

What did the federation ask Meta to do?

Reply within 30 days with practical steps to improve detection and prevention of impersonating ads, and set up a fast channel for businesses with dedicated resources, direct impersonation reporting, initial review within a few hours and immediate action on suspicious ads when there are real signs of fraud (ynet).

How did Meta respond?

Facebook Israel said it takes scams seriously, removed more than 159 million scam ads last year, wants to work with the federation and Israeli businesses, and rejects the claim that it accepts this activity (ynet). In March 2026 Meta also said it would expand advertiser verification (WIRED).

How does this relate to Bank Leumi's lawsuit against Meta?

It is a separate action. Leumi sued Meta in August over ads impersonating the bank (Calcalist). The federation's letter cites the suit as an example without taking a position on it, and Leumi's chief executive publicly backed the federation's complaint (ice).

What can a brand do while a fake ad is still running?

Report the ad and the page through the platform's brand channels, pursue the landing domain with its registrar and host, keep dated evidence of both, and warn customers with specific details. Meta says businesses can report scam ads that misuse their brand name through Brand Rights Protection (Press Gazette). Our guide to cloned and malicious ads covers the steps.

Sources