Md. Himel Hasnat Rafi

Fellow, Dismislab
Spyware attacks through Ads without any click: What is ‘AdFection’?

Spyware attacks through Ads without any click: What is 'AdFection'?

Md. Himel Hasnat Rafi

Fellow, Dismislab

Your device might be exposed to spyware without you even clicking any link or opening any file. In 2024 and 2025, spyware attacks targeted at least 45 devices belonging to Al Jazeera, the Qatar-based news outlet, through online advertisements. In this kind of attack, the user does not have to click anything, rather the attack begins as soon as the ad loads on the device, even if the ad isn’t visible on the user’s screen.

Named “AdFection,” the technique turns an ordinary online ad into a potential tool for surveillance, raising concerns about digital privacy and the security of the advertising ecosystem.

On October 15, Citizen Lab, a research institute based at the University of Toronto in Canada, published a report on the matter. The researchers said the Israel-based spyware maker Candiru (which later changed its name to “Integrity Labs”) was involved in the attack on Al Jazeera’s devices. According to the Citizen Lab report, this is the first public technical evidence of a zero-click mercenary spyware attack carried out through targeted online ads. They named the technique “AdFection,” which combines the words “ad” and “infection.”

Although some reports had earlier been published on the link between mercenary or for-hire spyware and the online advertising system, how such spyware makers exploit the advertising system to target specific individuals had remained largely unknown until now. Citizen Lab’s research presents technical evidence of zero-click spyware being delivered to a device through targeted ads. According to the researchers, in this method, an ad created to target a specific person can be enough to infect the device with spyware once it loads.

The research says the ads carrying the spyware were purchased by the Israel-based company Discover-Tech. Although the company presented itself as a legitimate advertising firm, in reality it acted as a link between the online advertising system and mercenary spyware companies.

How it works

In 2024, Al Jazeera’s own technology team detected some unusual communication patterns on devices used by several journalists. In 2025, Al Jazeera gave the network logs from that period to Citizen Lab. By analyzing these logs, the researchers obtained a step-by-step picture of how the attack was carried out.

The network logs of the infected devices show that they first communicated with domains under Discover-Tech’s infrastructure. This indicates that an ad was displayed on the device. The devices were then seen connecting to a Candiru validator server. This server performs an initial verification of the device. After that, more information about the device is collected and the spyware infection process advances further. Finally, the infected device communicates with Candiru’s command-and-control server.

Discover-Tech is the link between advertising platforms, including Google, Yandex, and Beeswax, and mercenary spyware companies like Candiru, Bindecy, and Intelos. Source: Citizen Lab.

Whenever an app or website has space for an ad, an automated auction takes place to decide which ad will appear. Advertisers bid for the opportunity to show their ads to a particular user. This process, known as real-time bidding, takes less than a second. The user’s device sends a request, advertisers place their bids, and an ad exchange selects the winner and delivers the ad to the app or website. 

Ad exchanges are digital marketplaces that facilitate these auctions and are often operated by companies such as Google. Hundreds of billions of auctions take place on these ad exchanges every day. 

Online ads are not just images or videos. They can also carry code provided by the advertiser, a DSP (Demand-Side Platform), a Supply-Side Platform (SSP) or another third party. Such code is generally called “tags,” “pixels,” or “SDKs (software development kits).”

When the ad is displayed, this code runs inside the app or browser. It is also used to collect information on the device’s identifier number and characteristics, location, IP address, which app or website is being used, and details of the ad space. The SSP sends these information to the ad exchanges for the purpose of auctions, making it possible to show ads targeted at a specific person.

Discover-Tech first wins an ad auction and delivers an ad to a website or app open on the user’s device. Once loaded, the ad requests JavaScript from a Discover-Tech server embedded within the ad. Servers for users who are not targets do not send back any code to execute. However, the ad still collects information from the user’s browser, including website cookies and the device’s public IPv6 (Internet Protocol version 6 ) address, and sends it back to Discover-Tech.

Whereas, for targets, Discover-Tech sends further JavaScript. In a later step, it installs spyware on the device by sending code that exploits a vulnerability in the browser. 

In the whole process, the user does not have to make a single click; all of it can be carried out in the background.

To learn more about this, Dismislab spoke with international digital security expert Bahaa Nasr. He said, “The ad appears on the screen. The JavaScript attached to it checks whether this is the device of the person they want to hack. If it is, it installs spyware on the device.”

In Citizen Lab’s words, “The display of a targeted ad is enough for infection.” In other words, even if the user does not have to click on a link or open a file, the spyware infection process can begin when a targeted ad loads on the device.

Ads on most platforms, even sensitive networks

Ads are displayed on numerous apps and websites. Citizen Lab’s analysis found that between March and June 2026, ads carrying the attackers’ script tags were shown to users of 75,373 Android and iOS apps. Similar ads were displayed on 120,768 websites in mobile and desktop browsers.

Discover-Tech ads can pose an infection risk. Researchers observed these while browsing the CBC and Yahoo websites, as well as while using a Sudoku app on Android. Source: Citizen Lab.

On Android, more than 1 million ads were shown on MX Player and Lark Player. More than 700,000 were shown on Viber. Nearly 500,000 ads each were shown on Snap VPN, Truecaller, and VivaCut. These ads were also seen in apps such as Yahoo Mail, Opera Mini, Waiter Online, Duolingo, WeMuslim, Period Calendar, Bubble Shooter, and Ludo Star. The number of ads on iOS was lower than on Android. For example, nearly 200,000 ads were shown on Viber on iOS, while on Android the number was more than 700,000.

These ads were also seen on websites. Of the auction bids analyzed, 23.3 percent were for displaying ads in web browsers. More than 1 million such ads were shown on the websites of several media outlets in Armenia and Serbia. These ads were also seen on the websites of the BBC, CBC, Haaretz, and The New York Times.

Citizen Lab’s analysis found that Discover-Tech’s ads were also displayed on networks tied to national security in various countries around the world. These networks include defense ministries, armed forces, national cybersecurity agencies, international and intergovernmental security agencies, and institutions connected to critical infrastructure. These ads were also shown on the networks of media outlets in various countries. The researchers also found evidence of ads being displayed on the networks of the United Nations, NATO, the European Parliament, the European Commission, and the General Secretariat of the Council of the European Union.

However, the research found no evidence that the owners of these apps or websites were involved in the attack. Nor does the display of such ads on an app or website mean that its users were infected with spyware. The researchers’ assessment is that, at the very least, through displaying these ads Discover-Tech gained the opportunity to collect data on many users and build profiles of them.

Are only journalists or officials at risk

Do ordinary people, apart from journalists or officials, have reason to worry about this? In response to this question, Bahaa Nasr said, “Everyone should think about this, because it is a question of their privacy. Everyone has to be aware.”

However, direct spyware attacks are expensive and are usually carried out against specific individuals. So for most people, the risk of this kind of attack is comparatively low. On the other hand, journalists, government officials, and people who work with sensitive information may be at higher risk.

Even before spyware infection, however, there is a risk of data collection and profiling of large numbers of people. The assessment of Citizen Lab’s researchers is that, as preparation for targeting specific individuals with spyware, data from millions of devices may have been collected and their distinctive characteristics identified. In other words, even if spyware does not enter a device, the advertising system creates an opportunity to collect information about the user’s identity and location.

With no click and no message, the person attacked may not notice the attack. Stephen O’Shaughnessy, a lecturer in cybersecurity at TU Dublin in Ireland, told Al Jazeera, “There is almost nothing left for the victim to notice.”

The attack also has a technical limitation, however. According to cybersecurity analyst Eddy Willems, an attacker does not gain control of the entire device just by being able to run some code inside the browser. To break out of the browser’s isolated protected environment, or “sandbox,” the attacker may have to exploit one more security vulnerability. In advanced spyware attacks, it is possible to use multiple vulnerabilities together.

Once spyware infects a device, it can be used to collect information such as files, documents, and the browser’s cookies, bookmarks, and saved passwords. It is also possible to record what the user types, take screenshots, capture audio and video through the microphone and camera, and collect emails and messages from various communication channels. To evade security systems, spyware often uses encrypted connections and sends stolen data in small pieces rather than all at once.

Advertising industry failed to catch it

From 2020 to 2026, Discover-Tech was on the lists of “approved” and “certified” ad vendors of Google and the advertising industry body IAB Europe. According to Citizen Lab, the company exploited this weakness in the advertising system. The research presents this case as an example of a serious lack of vetting in the digital advertising industry. In the researchers’ view, too much reliance was placed on information provided by the ad vendors themselves when they were added to these lists, and there was a lack of effective verification and accountability.

Before the research findings were published, Google and IAB Europe removed Discover-Tech from their ad vendor lists. Google shut off the company’s access to its ad exchange system. Beeswax removed Discover-Tech from its ad-buying platform. Vungle (Liftoff) temporarily suspended the company’s access and began its own review. Eskimi blocked all of Discover-Tech’s accounts and related domains.

In 2021, the US Department of Commerce added Candiru to the “Entity List.” According to the department, the company had developed spyware that was used to target government officials, journalists, businesspeople, activists, academics, and embassy staff.

Bangladesh context

Citizen Lab’s research found evidence of Discover-Tech ads being displayed on government-run or government-controlled networks in 24 countries. Bangladesh is not on this list. Neighboring Bhutan, Pakistan, Indonesia, Malaysia, and Thailand are. The list was prepared based on data from a limited period. So the absence of Bangladesh from the list cannot be seen as proof that there is no risk of such ads or attacks. Likewise, the display of ads on a government network does not necessarily mean that a spyware infection occurred there.

According to the website of Eskimi, the Lithuania-based digital advertising company, it has an office in Bangladesh as well. The company told Citizen Lab that from 2022, Discover-Tech ran ads on its ad-buying platform under its own management. Eskimi claims that because it blocked their campaigns and ad tags, only a few hundred dollars were spent on advertising by Discover-Tech. The company’s last ad campaign was shut down on August 6, 2026. Later, all of Discover-Tech’s accounts, users, and related domains were permanently blocked.

However, there is no evidence to conclude, from Eskimi having an office in Bangladesh or from Discover-Tech running ads on its platform, that any attack took place in Bangladesh or that Eskimi’s local office was involved in it. Citizen Lab’s research also found no such information.

In the interview, Bahaa Nasr said that journalists, human rights activists, political figures, and researchers who work with sensitive information in Bangladesh may be at high risk of this kind of targeted surveillance. However, no evidence has been found that such an attack has taken place in Bangladesh.

What to do to protect yourself from AdFection

With no click and no message, the person attacked may not notice the attack. Stephen O’Shaughnessy, a lecturer in cybersecurity at TU Dublin in Ireland, told Al Jazeera, “There is almost nothing left for the victim to notice.”

However, the attack has a technical limitation. According to cybersecurity analyst Eddy Willems, an attacker does not gain control of the entire device just by being able to run some code inside the browser. To break out of the browser’s isolated protected environment, or “sandbox,” the attacker may have to exploit one more security vulnerability. In advanced spyware attacks, it is possible to use multiple vulnerabilities together.

Once spyware infects a device, it can be used to collect information such as files, documents, and the browser’s cookies, bookmarks, and saved passwords. It is also possible to record what the user types, take screenshots, capture audio and video through the microphone and camera, and collect emails and messages from various communication channels. To evade security systems, spyware often uses encrypted connections and sends stolen data in small pieces rather than all at once.

In cases like these,  Citizen Lab warns that avoiding suspicious links or files is not enough to prevent attacks like AdFection, as spyware is installed through ads without any active step by the user. According to the institution, there is still no reliable way to keep an individual’s or an organization’s devices and networks completely safe from the risk of targeted ads. Some steps can reduce the risk, however.

  1. Use ad and tracker blockers. Bahaa Nasr suggested using an open-source extension called uBlock Origin Lite. It can be used if your browser supports it. Using Privacy Badger as well can help reduce online tracking. Before installing an extension, however, it is important to check its source and reliability.
  2. Regularly reset or delete your phone’s advertising ID. On Android, it is found in the privacy or ads settings. On iPhone, you can keep app tracking permission turned off. The names or locations of the settings may differ depending on the phone model and operating system version.
  3. Use personalized ad controls. Google’s and Facebook’s ad settings offer the option to limit or turn off personalized ads. Bahaa Nasr has kept such ads turned off on his own Google account. However, turning off personalized ads does not completely stop data collection or targeted ad display through other ad networks.
  4. Use browsers that offer extra privacy protection. Browsers like Brave offer the ability to block ads and trackers. However, no browser guarantees that it can stop all kinds of ads or surveillance.
  5. Keep your phone, browser, and apps regularly updated. Use the updated versions of software. Also be careful when giving out personal information online, and avoid sharing more information than necessary.
  6. Block ads on an organization’s or newsroom’s network. Measures can be taken to block ad and tracking connections not just on personal devices but across an organization’s entire network. Bahaa Nasr suggested running Pi-hole on a Raspberry Pi. Pi-hole is a DNS-based system that can block connections to many ad- and tracker-related domains for the various devices on a network. It cannot block all ads. Citizen Lab has also recommended blocking ads at the entry point of an organization’s network and on the devices used for work.
  7. If you are at high risk, take extra security measures. Those who are at risk of targeted surveillance, including journalists, can keep iPhone’s Lockdown Mode or Android’s Advanced Protection turned on. According to Citizen Lab, if someone’s work or identity puts them at higher risk, they should get help from an expert to have the security of their devices and network checked.

Is an ad blocker enough? In response to this question, Bahaa Nasr said, “It is the simplest measure, because the ads will simply not be shown anymore. But hackers are always coming up with new techniques.”

The Al Jazeera case has shown that it is possible to deliver spyware to a device using targeted online ads, without any click or file being opened. It is also difficult for the user to detect the attack. Most people may not be direct targets of this kind of attack. However, the online advertising system offers opportunities to collect and analyze various pieces of user information, which can be used to identify and target specific individuals.

So measures such as using ad blockers, controlling the advertising ID, and keeping the phone, browser, and apps updated should be considered an important part of digital security. But these measures can reduce the risk; they cannot eliminate it. To prevent attacks of this kind, it is also essential to ensure effective vetting, security monitoring, and accountability by advertising industry institutions.