Zero-Click Spyware Ads: Citizen Lab's AdFection Report

Citizen Lab published a report on October 15, 2026 describing a technique it says governments have used to hack targets through ordinary online advertising, with no click, no download, and no warning. The group calls the technique AdFection , and says it marks the first public confirmation that zero-click mercenary spyware has been delivered through targeted ads. The story was first flagged widely after TechCrunch covered the report, and it lands at a moment when the ad-tech industry is already under pressure over data collection and targeting practices.

At the center of the report is an Israeli advertising company, an ad platform most people have never heard of, and two commercial spyware makers. Below is what Citizen Lab confirmed, what remains unverified, and why researchers are calling this a turning point for the surveillance industry.

Citizen Lab Exposes “AdFection”: What the Report Says

The report is titled “Exposing AdFection: How Discover-Tech Turned Targeted Ads Into a Pipeline for Zero-Click Mercenary Spyware,” and it was published by the Citizen Lab research group at the University of Toronto’s Munk School. AdFection is the name Citizen Lab gave to the method itself: using targeted online advertising as a delivery pipeline for spyware that activates without any action from the person being targeted.

Citizen Lab described the findings as “the first-ever public confirmation of zero-click mercenary spyware delivered via targeted ads.” That framing matters. Commercial spyware has existed for more than a decade, but researchers have mostly traced infections to text messages, malicious links, or network-level attacks. An ad slot that silently triggers an infection the moment it renders is a different threat model, and a much harder one to see coming.

How Zero-Click Spyware Hides Inside Targeted Ads

According to the report, a targeted individual did not need to click a link, open a message, or download a file to be compromised. The attack could trigger simply when a specific ad was displayed on the person’s device. That single fact is what separates AdFection from almost every other spyware delivery method documented over the past decade.

Modern online advertising runs through a chain of automated auctions that happen in the time it takes a page to load. A simplified version of that chain looks like this:

That automated, high-speed handoff between ad exchanges, publishers, and ad networks is exactly what makes targeted advertising attractive as a surveillance pipeline. An attacker who can get a malicious creative into the bidding pool, and who can target a specific device or user profile, does not need the victim to do anything at all.

Citizen Lab identified Discover-Tech, an Israeli advertising company, as the firm behind the pipeline. Discover-Tech’s advertising platform is called Controlium, and Citizen Lab said at least two zero-click spyware makers, Candiru and Bindecy, used Discover-Tech’s platform to reach their targets.

Candiru and Bindecy: The Spyware Makers Behind AdFection

Candiru is a name familiar to anyone who has followed the commercial spyware industry. The Israeli firm has shown up in prior research tied to surveillance tools sold to government clients. Bindecy is a far less documented name, and this report is one of the more prominent public appearances of that vendor to date.

Citizen Lab also linked Candiru spyware specifically to the attacks it documented at Al Jazeera. The involvement of two separate spyware vendors on a single ad platform suggests Controlium functioned less like a single client’s tool and more like shared infrastructure, a kind of marketplace where different surveillance vendors could plug in.

Al Jazeera: 45 Devices Targeted Across Two Years

Citizen Lab reported that 45 devices at Al Jazeera were targeted or compromised during 2024 and 2025. The affected devices were reported as Windows machines, according to NBC News , which also detailed the Israeli ad company’s alleged role in helping a spyware firm reach journalists through online ads.

Al Jazeera has been a recurring target of commercial spyware for years, showing up in earlier Citizen Lab research tied to other surveillance vendors. A news organization with staff spread across conflict zones and politically sensitive beats is, unfortunately, a predictable target for any government looking to monitor journalists, sources, or leaked information before it reaches publication.

Governments, NATO, and the UN in the Blast Radius

Al Jazeera was not the only entity exposed to the malicious ads. Citizen Lab said the ads were also shown to people at the United Nations, NATO, and the European Commission, as well as at more than a dozen national governments. That list reads less like a narrow journalist-tracking operation and more like a broad surveillance net cast across international institutions.

Citizen Lab has not identified which governments were behind the attacks. That detail remains unconfirmed, and readers should treat any claim naming a specific government as speculation until Citizen Lab or another credible investigator publishes attribution evidence.

Confiant and the Investigation That Cracked the Case

Citizen Lab did not work alone. Confiant, a company that specializes in detecting malicious advertising, participated in the research alongside Citizen Lab. Ad-fraud and malvertising researchers like Confiant spend their days hunting for malicious creatives hiding inside legitimate ad exchanges, which made the firm a natural partner for a story that lives at the intersection of ad-tech and spyware.

That partnership is also a signal to the ad-tech industry. Malvertising, the practice of smuggling malware through legitimate ad networks, has existed for years in cruder forms, usually tied to scams, cryptocurrency theft, or browser hijacking. AdFection shows the same delivery mechanism scaled up to nation-state-grade spyware.

Historical Context: From Pegasus to Programmatic Ads

Commercial spyware has followed a rough evolution over the past ten years. Early tools leaned on spearphishing links sent by text or email, hoping a target would tap once. Later generations moved toward zero-click exploits buried in messaging apps, where a crafted file or message could trigger an infection the moment the device processed it, without the user ever opening anything.

What those earlier zero-click methods had in common was a dependence on a specific app or protocol: a messaging client, an operating system component, a particular file format. AdFection breaks that pattern by using the advertising ecosystem itself, something nearly every connected device touches constantly, as the delivery layer. Targeting shifts from “which messaging app does this person use” to “which ad network serves this person’s device,” a far broader and harder-to-patch surface.

That shift also changes who has to respond. A messaging-app exploit is ultimately a problem for one company to patch. An ad-delivered exploit touches every link in the programmatic chain: the exchange running the auction, the demand-side platform placing the bid, the publisher rendering the creative, and the browser or app executing whatever code comes with it. No single company can patch its way out of that alone, which is part of why Citizen Lab and Confiant framed this as an industry-wide finding rather than a single-vendor bug.

Key Facts at a Glance

Spyware Delivery Vectors Compared

AdFection does not replace older spyware delivery methods, it adds a new one. The table below lines up the delivery method documented in this report against the categories of attack commercial spyware vendors have relied on in the past, based on general public reporting on the surveillance industry.

The ad-tech industry already fights a credibility problem. Advertisers worry fraud, publishers worry brand safety, and regulators in the EU and US keep tightening rules around targeting and consent. AdFection adds a new worry to that list: the possibility that the same real-time bidding infrastructure powering a shoe ad could, in rare and highly targeted cases, carry a nation-state payload.

That is a very different scale of problem than a typical corporate data breach . A breach like the one that exposed code and personal data tied to roughly 1,500 people in the Meta GitHub breach is bad for the people affected, but it is contained to stolen records. AdFection describes an active delivery mechanism that can reach anyone an attacker chooses to target, as long as that person’s device sees the right ad.

For ad exchanges and demand-side platforms, the pressure now shifts toward auditing who is allowed to buy targeted inventory and how creatives get vetted before they serve. For enterprises, the report is a reminder that endpoint security and network monitoring, the usual focus of teams tracking issues like the Ivanti VPN zero-day , now has to account for the browser and the ad slot as an attack surface too.

Publishers face a quieter but real exposure too. A news site, a weather app, or a productivity tool that simply runs third-party ad code has no easy way to inspect every creative that wins an auction in real time. That lack of visibility is exactly why malvertising has survived as a category of attack for so long, and why Confiant and similar firms built entire businesses around scanning ad creatives before they reach a reader’s screen. AdFection raises the stakes for that scanning work from catching scams and cryptocurrency drainers to catching state-grade spyware.

What Experts Are Saying

Bill Marczak, a senior researcher at Citizen Lab, described the scope of what AdFection enables. “AdFection allows operators to automatically hack specific individuals with mercenary spyware as they use apps or browse the internet,” Marczak said, according to Wired .

Marczak also described how long Discover-Tech had reportedly operated in this space. “Discover-Tech has leveraged the advertising ecosystem for years to turn trusted websites and popular apps into a zero-click spyware delivery platform,” Marczak said, according to Engadget. Both point to the same underlying concern: this was not a one-off exploit, it was infrastructure built to be reused across targets and, apparently, across spyware vendors.

What We Still Don’t Know

A report this detailed still leaves real gaps. Citizen Lab has not named the governments behind the campaign, so any specific accusation circulating online should be read skeptically. The organization also has not said that every ad served through Controlium carried a spyware payload, since tracking and profiling appear to have been separate, more common uses of the platform.

It also is not yet clear how many organizations beyond Al Jazeera were actually compromised rather than simply shown the ads. Being served a malicious ad is not the same as being infected, and Citizen Lab’s reporting draws that line carefully even where outlets covering the story have sometimes blurred it.

Predictions: Where This Goes From Here

Expect ad exchanges and demand-side platforms to tighten vetting for political and government-adjacent ad buyers, similar to how election-ad rules evolved after past controversies.

Expect at least one lawmaker or regulator in the EU or US to cite AdFection directly when pushing new rules on ad-tech transparency and targeting data.

Expect other security research groups to go looking for similar ad-delivered payloads on different platforms, since Citizen Lab’s method for spotting this one is now public.

Expect Discover-Tech, Candiru, and Bindecy to face renewed scrutiny from export-control regulators, echoing the pattern seen with earlier commercial spyware vendors.

Expect browser makers and antivirus vendors to add detection signatures aimed specifically at ad-delivered payloads rather than only link- or file-based threats.

What Advertisers and Users Can Do Now

There is no simple setting that blocks a zero-click ad-delivered exploit, which is part of what makes this report unsettling. Keeping operating systems and browsers fully patched remains the strongest practical defense, since most spyware ultimately relies on an unpatched flaw to execute once the ad renders.

Journalists, activists, and government staff in sensitive roles are the most realistic targets for a tool like this, and organizations supporting that kind of work should treat endpoint detection and device compartmentalization as a baseline, not an extra. Broader guidance on securing everyday AI and productivity tools against data exposure, covered in our piece on AI assistant security and your data , applies here too: assume any connected device is a potential target and reduce what it can access accordingly.

The broader pattern also matters. State-linked actors reaching targets through infrastructure most people trust by default, whether that is an ad network or, as seen in reporting on North Korea’s WaterPlum campaign , cryptocurrency platforms, keeps showing up across very different sectors. The common thread is attackers finding the systems that everyone already uses and quietly turning them into a weapon.

Frequently Asked Questions

AdFection is the name Citizen Lab gave to a technique for delivering zero-click mercenary spyware through targeted online advertising, documented in a report published October 15, 2026.

Who is Discover-Tech?

Discover-Tech is the Israeli advertising company that Citizen Lab identified as operating the ad platform, called Controlium, used to deliver the spyware.

Which spyware companies were involved?

Citizen Lab said at least two zero-click spyware makers, Candiru and Bindecy, used Discover-Tech’s Controlium platform to reach targets.

Was Al Jazeera the only organization targeted?

No. While 45 devices at Al Jazeera were confirmed targeted or compromised in 2024 and 2025, Citizen Lab said the malicious ads were also shown to people at the United Nations, NATO, the European Commission, and more than a dozen national governments.

Which governments were responsible for the attacks?

Citizen Lab has not publicly identified which governments were behind the attacks. That remains unconfirmed as of this report.

Keep operating systems, browsers, and apps fully updated, since zero-click exploits generally rely on unpatched software flaws. People in high-risk roles, such as journalists and activists, should also consider device compartmentalization and specialized endpoint monitoring.