Security issues and browser warnings
Three families trigger a warning: malware, unwanted software and social engineering. The security report separates what operates without any visitor action from what the visitor downloads themselves. And deceptive content embedded in an advertisement is a policy violation for the page hosting it.
A mechanism separate from everything else
This site devotes several guides to what moves a page up or down. The subject here is not one of them: it concerns a warning shown before the site opens, by the browser, to protect the person about to click. The mechanism, the report and the remedy all differ from those of a ranking penalty.
Malware is software or a mobile application specifically designed to harm a device, the software it runs or its users; the documentation notes that site owners sometimes host such files inadvertently.
officialSource : Google Search Central, “Malware and unwanted software”
The second half of that sentence deserves a pause: site owners sometimes host these files without knowing. That is not a courtesy, it is a description of the commonest case — a compromised site, an outdated module, an upload directory left open.
Unwanted software is an executable file or mobile application whose behaviour is deceptive, unexpected, or degrades the browsing experience: the documentation cites switching the home page without consent and leaking personal information without proper disclosure.
officialSource : Google Search Central, “Malware and unwanted software”
The distinction is useful because it moves the criterion: it is no longer about harming, but about doing something other than what was announced. A perfectly functional program that changes a setting without saying so falls into this category.
In the Security Issues report, “Malware” refers to what operates without explicit user action, while “Harmful downloads” refers to what the user must explicitly download.
officialSource : Google Search Central, “Malware and unwanted software”
What social engineering covers
Social engineering is content that tricks a visitor into doing something dangerous: either it pretends to look and feel like a trusted entity — browser, operating system, bank, government — or it obtains an action one would only take for such an entity, such as sharing a password or calling support.
officialSource : Google Search Central, “Social engineering (phishing and deceptive sites)”
Two criteria, then, and the second is wider than it looks: obtaining an action one would only take for a trusted entity. A window announcing that the system is out of date and offering an update fits that description exactly, even if nothing gets installed.
A provider operating a service on behalf of another party without making that relationship clear may be flagged as social engineering; the example given is a donation platform acting for a charity.
officialSource : Google Search Central, “Social engineering (phishing and deceptive sites)”
That third case is the most surprising for an honest site. A provider collecting donations, booking appointments or handling a form on behalf of somebody else must state clearly that it acts for that party. Otherwise the page looks like something it is not, and the resemblance is enough.
The rule that makes you answerable for what you did not write
Social engineering content embedded in a page, usually through an advertisement, is a policy violation for the host page.
officialSource : Google Search Central, “Social engineering (phishing and deceptive sites)”
This is the point to take away from the whole guide. A site displaying ads chosen by a third party remains answerable for what appears on it. The documentation adds that such content is not always visible from the publisher’s own view of the page, which makes checking harder and the flag more surprising.
An advertisement that does not state what will be downloaded is covered: the documentation cites an ad carrying only the words “Download” or “Play” without identifying the software, and a play button that leads to a download.
officialSource : Google Search Central, “Malware and unwanted software”
It is not our content: the banner comes from the ad network. We have passed it on, they will handle it, the site has nothing to do with it.
Watching the security report and what the ads actually display
hypothesis
- Effort
- moderate
- Time to effect
- immediate
- Risk
- none
- Durability
- lasting
The effort lies in regularity rather than difficulty. The delay is immediate for the finding, longer for recovery, which goes through a review request. The risk is nil, monitoring not being a technique. Durability is medium: modules, ad networks and third-party scripts change, and the check has to be redone.
What does not work, or no longer works
Passing responsibility to the ad network. The documentation writes that embedded content binds the page hosting it.
Then treating a security warning as a ranking penalty. The two mechanisms have different reports, different causes and different remedies.
Finally, assuming a clean site cannot be concerned. The documentation itself notes that files are sometimes hosted without the owner knowing, which is the definition of a compromised site.
Sources
- Malware and unwanted software — Google Search Centralaccessed on
- Social engineering (phishing and deceptive sites) — Google Search Centralaccessed on
Block revalidated on
Frequently asked questions
- Is a security issue the same as a manual action?
- No, and the two live in different console reports. A security issue protects visitors and can trigger a warning the browser shows before the site even opens. A manual action penalises breaches of the spam policies and acts on position instead.
- Can an ad network get you flagged?
- Yes, and it is the least known rule here: deceptive content embedded in a page, usually through an advertisement, is a policy violation for the host page. The site remains answerable for what it displays, including what it did not choose line by line.
- What separates malware from unwanted software?
- The first is designed to harm, the second merely behaves deceptively or unexpectedly: switching the browser home page, leaking personal information without saying so. The security report additionally separates what runs without user action from what the user downloads themselves.
- Can a download button be a problem?
- The documentation names two cases: an ad carrying only the words “Download” or “Play” without identifying the software, and a play button that actually triggers a download. The criterion is the gap between what the button announces and what it does.
- What do you do when a warning appears?
- Open the Security Issues report, which lists the suspected files and the type of violation found. Once the problem is fixed, a review request is filed from that same report. Recovery depends on that review, not on an automatic delay.