Israeli Ad Tech Firm Uses Online Ads to Deliver Spyware via Zero-Click Attacks

According to a joint investigation by security researchers at Citizen Lab and Confiant, an Israeli ad tech company has partnered with a spyware vendor to implant malicious programs on user devices through online ads. Targets include devices belonging to individuals associated with Al Jazeera. This marks the first time researchers have found technical evidence of a zero-click attack — users need not click any link or download any file; simply browsing a webpage with the ad content may be enough to compromise their device.

Ad Poisoning: From Marketing Tool to Attack Vector

Traditionally, online advertising is seen as a tool for brand promotion and traffic monetization. But this investigation reveals a radically different abuse path: the distribution power of ad networks has been weaponized into a vehicle for precisely delivering spyware.

The core of the attack chain lies in the automation and broad reach of ad delivery. Ad tech platforms are typically connected to numerous websites, ad exchanges, and data analytics services, capable of pushing content to targeted audiences worldwide within seconds. When this infrastructure is maliciously exploited, attackers need not directly touch the target device — simply embedding attack code into the ad supply chain allows them to compromise users through normal web browsing behavior.

According to the investigation, the ad tech company involved had a partnership with a spyware vendor, packaging the latter's attack payload as ad creative and delivering it through ad exchanges to websites the targets were likely to visit. In Al Jazeera's case, members of its audience could become infected simply by browsing pages containing the specific ads.

Why Zero-Click Attacks Are More Dangerous

A zero-click attack means the attack process requires no interaction from the victim whatsoever. Phishing emails require users to click a link; malicious downloads require users to actively execute a file. But zero-click attacks rely solely on software vulnerabilities triggered when a device processes specific content — the trigger process is completely transparent to the user.

This attack method is especially dangerous for high-value targets. Journalists, human rights activists, political dissidents, and similar groups typically have some security awareness and avoid clicking suspicious links or downloading unknown attachments. Zero-click attacks bypass these defensive habits entirely, exposing even the most cautious users to risk. Al Jazeera, as a major media organization in investigative journalism, has long faced various cyber threats against its editorial staff. Being named as an attack target further confirms the stealth and penetration power of ad-based delivery channels.

Researchers noted that this is the first time technical evidence has been obtained showing a zero-click attack delivered via online advertising. While there had been speculation and theoretical discussions in the industry previously, verifiable samples and attribution analysis were lacking. The joint investigation by Citizen Lab and Confiant fills this gap.

Zero-click attacks typically exploit unpatched vulnerabilities in operating systems or application software — particularly memory corruption flaws in browser rendering engines, image processing libraries, or message parsing components. Attackers craft malicious ad creative (such as specially constructed images, fonts, or scripts) to trigger these vulnerabilities, achieving remote code execution on the device. Since the entire process requires no user interaction, traditional security prompts and warning mechanisms based on user behavior become completely ineffective. Moreover, zero-click exploit chains often involve combinations of multiple vulnerabilities, increasing the complexity and cost of the attack — which is why they are typically used for targeted attacks against high-value targets. In recent years, disclosed zero-click attack cases have included exploits targeting iMessage, WhatsApp, and other instant messaging apps, but zero-click attacks via online advertising have never before been technically confirmed — until now.

The Systemic Security Dilemma of the Ad Ecosystem

This incident exposes deep-seated security issues in the ad tech supply chain. From advertiser to final display on a user's screen, an ad typically passes through multiple stages: demand-side platforms (DSPs), supply-side platforms (SSPs), ad exchanges, data management platforms, and more. Each stage involves the embedding and execution of third-party code, and these codes often lack rigorous auditing mechanisms.

The openness of ad networks makes it relatively easy for malicious actors to purchase ad slots and deliver content. Even if platforms have review processes, facing massive volumes of ad creative and constantly evolving attack techniques, both manual and automated reviews struggle to be foolproof. When spyware is carefully disguised as normal ad creative, detection and interception become even harder.

The impact extends beyond direct targets. Ad delivery is based on user profiling and behavioral targeting — theoretically, any user matching specific criteria could be included in the attack scope. This means ad-delivered attacks have inherent potential for scale; once abused, the impact could far exceed what any single targeted organization might expect.

Ad exchanges are the core hub of programmatic advertising, connecting demand-side platforms (DSPs) and supply-side platforms (SSPs), completing ad slot transactions in milliseconds through real-time bidding (RTB). In the RTB process, ad creative is typically embedded as HTML, JavaScript, or images — content often dynamically provided by third-party servers and potentially passing through multiple intermediaries. This highly automated and fragmented chain makes security auditing extremely difficult: a lapse at any single point could be exploited by attackers to inject malicious code. Additionally, ad networks typically allow advertisers to upload creative themselves, and automated review systems mainly target obviously violating content (such as pornography or violence), making it hard to identify obfuscated spyware payloads. Attackers can also leverage ad targeting features to precisely deliver malicious ads to specific geographic regions, device types, or user profile segments, further reducing the probability of detection.

Next Steps for Security Defense

For ordinary users, the means to counter such attacks are relatively limited. Keeping operating systems and browsers updated to the latest versions can somewhat reduce the window for known vulnerabilities to be exploited. Using browser extensions or security tools with ad-blocking capabilities can lower the probability of exposure to malicious ads but cannot completely eliminate the risk.

For enterprise users and news organizations, this incident suggests the need to re-examine security strategies around web browsing. Network isolation, traffic detection, and endpoint protection need to incorporate the ad supply chain into their threat models, rather than focusing solely on traditional phishing and malicious download channels. For highly sensitive positions, using isolated browsing environments or disabling third-party ad scripts may be necessary mitigation measures.

At the industry level, information-sharing mechanisms between ad tech companies and security vendors still need improvement. When researchers discover evidence of malicious ad delivery, there is currently a lack of standardized response processes for quickly identifying the ad networks involved, notifying potential victims, and disrupting the attack chain. The joint investigation by Citizen Lab and Confiant provides a collaborative model, but there is still considerable distance to go before industry standards are established.

Key Takeaways

• A joint investigation by Citizen Lab and Confiant found that an Israeli ad tech company partnered with a spyware vendor to deliver malicious programs to target devices through online ads.
• This is the first time researchers have found technical evidence of a zero-click attack delivered via online advertising — users can be infected without any interaction.
• Targets include individuals associated with Al Jazeera, indicating that journalists and media organizations are high-value target groups.
• The multi-stage embedding of third-party code in the ad tech supply chain provides a covert channel for malicious delivery, and existing review mechanisms struggle to effectively intercept it.
• Defense requires incorporating the ad supply chain into security threat models, and the industry needs to establish more robust malicious ad response and information-sharing mechanisms.