What a «This site may be dangerous» warning really means
A dangerous-site warning means the protection service has detected a possible risk to visitors. In Chrome, these alerts can come from Google Safe Browsing. They differ from “Your connection is not private”, which concerns the secure connection. Save the exact wording and check Search Console’s security report before deciding what needs fixing.
The warning means Google has found a specific signal on one or more pages of your site. It does not mean, by itself, that your server has been compromised, that your customers’ data is exposed or that the whole site is down. It can happen that an attacker injected code on a single URL, that a file uploaded months ago now triggers detection or, less often, that the warning responds to a misclassification. Before reacting blindly, it helps to know exactly what Google saw and on which pages.
Where the warning appears and how you learn about it
The warning does not live in a single place. It usually shows up on several channels at once and it is worth telling them apart because they do not say the same thing:
- In the visitor’s browser. Chrome may display a dangerous-site or deceptive-site warning. Other browsers have their own protections and the wording can differ. Chrome’s help page explains its warnings; do not disable protection to get past one.
- In search results. A warning may appear below the result title, something like «This site may harm your computer».
- In Search Console messages. You may receive a security notification. Open the dashboard through a trusted address and check the report there rather than relying on the email alone.
- A customer writing to you. Sometimes the visitor warning reaches an inbox before yours. That message is useful information: ask for a screenshot, time and device if they did not include them.
This matters because the browser warning targets the visitor and the Search Console report targets you. The first one protects whoever opens your site; the second is your main source for finding out which pages are flagged and which type of issue Google detected.
The categories inside the Security Issues report
The Security Issues report in Search Console groups what Google detects into a few categories. They are not equivalent, and reading the exact label saves time:
| Label in Search Console | What Google noticed | Where to look first |
|---|---|---|
| Hacked content | Pages with text, links or scripts that, according to Google, are not yours. | Recently modified theme and CMS files, unknown users with high privileges. |
| Malware | Code on the page that downloads or runs something harmful on the visitor’s device. | JavaScript loaded from unusual external domains and recently uploaded files. |
| Unwanted software | Covert installs or browser changes after visiting the page. | Buttons or ads nudging downloads, third-party scripts that are little known. |
| Social engineering | Pages impersonating a brand, a bank or a legitimate form to request information. | Forms or pages you did not create that impersonate another service. |
| Uncommon downloads | Files offered for download that match new or suspicious patterns. | Download area, attachments in recent posts or uploads from the admin panel. |
The report may show example affected URLs. Save those examples to support the investigation, but do not treat them as a complete inventory. The technical review needs to check whether the same issue affects other pages.
Before you change anything: save what you can see
Reacting fast is not the same as reacting without thinking. Before changing passwords, restoring a backup or deleting the file that looks odd, preserve information you will not be able to recover later:
- Take screenshots of the browser warning: the full message, the detail Google provides and the exact URL in the address bar.
- Keep the Search Console email with its full headers. If several arrived, store them by date.
- Note the list of affected URLs shown in the report, with the type of issue linked to each one.
- Record the browser, device and time you saw the warning from. If it reached you through another channel (customer, social media), note who alerted you.
- If you plan to change anything on the site, ask your technical team or hosting support for a copy of the current state (files and database) before starting.
If you also see content you did not publish or unexpected redirects, contact the team that maintains the site. Preserve available information without delaying measures needed to protect visitors. Our security assessment page explains our approach and its scope.
How to check the status from outside and from inside
You can double-check what is happening with your site from two different places: a public source and your Search Console panel. Each provides different information and it helps to look at them in order.
- Google Transparency Report. The public Safe Browsing lookup lets you enter a domain and tells you whether it currently appears as dangerous according to Google’s latest crawl. It does not include URL-level detail or the cause, but it shows the global state as the service perceives it.
- Search Console → Security Issues. Read each issue and save any example pages it provides. You can request a review there once the problem is fixed across the website, not just on the sample URLs.
- Do not bypass a warning to open the page. You do not need to disable protection or try several devices to confirm it. The technical team can examine the content in an appropriately prepared environment and use Google’s reports.
What not to do: treat your website as fine after a single check from your own device. A browser extension, a hosting cache or an open session can make you see something different from what a fresh visitor arriving from Google sees.
What this warning is telling you and what it is not
Understanding the scope of the warning avoids disproportionate reactions. As a rough summary:
- The URLs shown in the report are starting points for investigation. They do not establish that every other page is unaffected.
- The warning can affect a subdomain without its parent being flagged, or the other way around. If you work with subdomains (for example, a customer area), check each one.
- The warning is not a vulnerability report or an audit. It will not tell you which plugin or which line of code is the root cause. If you already have a preventive security report, cross-reference both; if not, the Google warning is a starting point, not a diagnosis.
- The warning does not imply your databases or your customers are compromised. It can be consistent with that, but also with a single file modified by a specific issue. Confirming it is a later step.
- A classification can also be wrong. That possibility does not justify ignoring a warning: the team should inspect the content and document why it considers the site clean.
With the scope in hand you can decide whether you are facing an incident, a configuration error or a routine review of the flagged content.
Coordinating the response with your team
An orderly response to a Safe Browsing warning usually involves three or four people, each with their own part:
- You, as the person responsible for the site. You decide what is communicated, whether to temporarily disable flagged pages and when the review is considered closed.
- Whoever maintains your website. They review the affected URLs, the recently modified files, the database and recent components. If they locate a pattern, they prepare the specific fragment that needs correcting.
- Hosting support. They can see complete server logs that neither you nor whoever maintains the website have from the admin panel. They are often the ones who spot suspicious access or unauthorised processes still running.
- Legal or data protection advice if personal data may be affected. This supports the responsible business in assessing duties and deadlines. A Google warning alone does not establish a personal data breach or replace that assessment.
Before restoring anything, check the backup’s date and what it includes. If the issue has gone undetected for a while, yesterday’s backup may be compromised just like the current site.
Request a review once the problem is fixed
Once you have corrected what needed correcting on the flagged URLs, you can request a review from the Security Issues report in Search Console. The procedure is described in the official help and essentially consists of marking the issue as resolved and adding a short explanation.
Three things to keep in mind before pressing the button:
- Request the review when the fix is real. If Google reviews and finds the issue again, the red screen will not only stay: it may extend the recovery time on subsequent attempts.
- Describe in a few lines what you changed. No forensic report is required: it is enough to state that you located the cause (injected file, replaced form, removed link), what you corrected and how it will be prevented from reappearing.
- Do not promise a date for the warning to disappear. Review times can vary. Keep the request confirmation and follow its outcome; do not send another request while one is pending.
If an intrusion is confirmed, also agree follow-up checks and an access review. Their duration depends on what happened and what the team can check; there is no single period that proves the website has recovered.
If you would like us to review it with you
If you need help understanding the warning, tell us the message you see and what Search Console reports. We can help establish which review may be needed and agree the work and timeframe before starting.
For a preventive assessment, see our security page. If you already have a technical document, this guide explains how to interpret its findings.