# Google "Safe" Browsing

This article was created and is maintained by Izzy.

About a month after the last Netcraft event, Google started playing a similar game – except for not even sending out reports (unless the domain was registered with their Webmasters). What do both seemingly have in common? Entirely relying on automation, no humans involved.

# Introduction

Google Safebrowsing / Webmasters seems to be a service which [even flags their own mails as suspicious](https://support.google.com/webmasters/thread/382326165?hl=en&msgid=382561704) so they get lost in the spam folder (at least if you use Gmail), and one wonders about not receiving reports, while the website is flagged as “potential dangerous” or “potential malicious”. So you have to register your website with them to see what they think to have found (no, they don't send you reports – you just hear from others that your site was obviously flagged) – and there you find the “report”. Here this started 2025-12-30:

# Event Log

## The first incident


 ![Website Screenshot from a visitor](https://docs.katastima.org/api/files.get?sig=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJrZXkiOiJ1cGxvYWRzL2E2ZTRmNWNhLWJjNTktNDNkZS1iZGJhLTdhNjljNWU1Y2NlZi9kYmI1OWUyOS0wNTY0LTQwZTQtOWM5Mi0zNmQ0ZTIzMTg2Y2QvaW1hZ2UucG5nIiwidHlwZSI6ImF0dGFjaG1lbnQiLCJpYXQiOjE3OTUyMzgxMDAsImV4cCI6MTc5NTMyNDUwMH0.SCKjCbwYp1DFVqNhqGAblrObc-PYvBUx1CSHRqWyd0o "left-50 =300x841") ![Website Screenshot from a visitor](https://docs.katastima.org/api/files.get?sig=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJrZXkiOiJ1cGxvYWRzL2E2ZTRmNWNhLWJjNTktNDNkZS1iZGJhLTdhNjljNWU1Y2NlZi8xZTlhYzg3My1mZTNkLTQxYzQtYmQ3NC00YzdmYzc2MmM2ZTQvaW1hZ2UucG5nIiwidHlwZSI6ImF0dGFjaG1lbnQiLCJpYXQiOjE3OTUyMzgxMDAsImV4cCI6MTc5NTMyNDUwMH0.Q_erGcakCfgD4T-SvF58csfHqG_HfG9BraXCE-CjrFA "left-50 =315x666")


#### 

I put “report” in quotes, as it is a joke: it says “1 problem found” – and the details say “not applicable” (and before you ask: the link to “further information” (in the screenshot below: “Learn more”) just goes to a general help page):


 ![The problem as shown in Google's console](https://docs.katastima.org/api/files.get?sig=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJrZXkiOiJ1cGxvYWRzL2E2ZTRmNWNhLWJjNTktNDNkZS1iZGJhLTdhNjljNWU1Y2NlZi9kODY2YjdlNS01NTM3LTQxNTAtYmMyMC00OGJmOThkYWRkYmMvaW1hZ2UucG5nIiwidHlwZSI6ImF0dGFjaG1lbnQiLCJpYXQiOjE3OTUyMzgxMDAsImV4cCI6MTc5NTMyNDUwMH0.uWqc6pIv4fSMGqrH6GA6AIsu7osAIYE5RBNBWvOKt0Y " =930x452")

They found “malicious/harmful content”, but do not know which (fun fact: checking the same site with VirusTotal reveals that Google marked it “clean”. Something with “left hand” and “right hand”?). I guess the “dangerous apps” are those present in our repository, but not listed at Google Play?

So I asked for reevaluation (there's a button for that: “Request Review” in the above screenshot), and they want to know how I solved the issue. Well, which issue? It simply left me to write I have “checked and cleared all indicated pages and files” (which is true: they indicated nothing). Surprise: the answer says reevaluation failed (“our systems still indicate that URLs on {your domain} contain content that can harm visitors”, but they still cannot give a single example; and no, we did **not** receive any mail about the verification either, just found that later in the notifications at their portal) – but the issue is cleared. Find the full story for this occasion [in this thread](https://floss.social/@IzzyOnDroid/115805947687179802).

## Next round

You’d think you’re done at that point – but nope. This is a repeating pattern. Just 20 days later (**2026-01-20**) we got another report of a "harmful file" they write they detected – while at the same time, at VirusTotal *their own engine* reports it as clean. [Details here](https://floss.social/@IzzyOnDroid/115927358849078868). Oh, and again: Just saw the report because I was checking open tabs in my browser. No mail received by neither admin of the domain there. Guess we won't receive any mail once their review is finished either…

**Update 2026-01-21:** Review failed. Obviously, not a single human involved (same template text as always). App in question is an APK Updater. Suspicious? Definitely! But not the app, the report and by whom…

## Groundhog day

**Update 2026-01-24:** [next one](https://infosec.exchange/@alcea/115949132103262807). And again, the file marked as “malicious” by safe browsing, is marked clean by their own scanner at VirusTotal ([link](https://www.virustotal.com/gui/file/966dd1ee9e2b83e5f8854c0520f794788adf461a918bfd4e545ffd79418946ac/detection/f-966dd1ee9e2b83e5f8854c0520f794788adf461a918bfd4e545ffd79418946ac-1756499291)).

**Update 2026-01-25:** Google's so-called “Safebrowsing” now **blocks our entire site:** each single APK is allegedly malware, see [our toot here](https://floss.social/@IzzyOnDroid/115955654662182732). Which is of course insane, as we scanned each single APK multiple times, and VirusTotal also confirms there are no findings (apart from the *occasional* Google Safebrowsing). This is not a safety service, it's Mafia behavior: “We rule the Internet, OBEY!!!” (or, “nice website you have there. Would be a pity if…”).

So as with Netcraft, so far only false positives, wasting our time needlessly.

## Other site, similar game

**2026-03-15:** today I received another “malware alert”:

 ![Screenshot from Google's console, showing allegedly malicious URL which does not even exist on the indicated website](https://docs.katastima.org/api/files.get?sig=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJrZXkiOiJ1cGxvYWRzL2E2ZTRmNWNhLWJjNTktNDNkZS1iZGJhLTdhNjljNWU1Y2NlZi9hZGU3M2ZhNC0yZDg4LTQ5ZGItOGJkOC03ZTE0MjBkZGY0Y2YvaW1hZ2UucG5nIiwidHlwZSI6ImF0dGFjaG1lbnQiLCJpYXQiOjE3OTUyMzgxMDAsImV4cCI6MTc5NTMyNDUwMH0.LywLNC-IR4kVXWBVbqqeV2fg2CTUdBRnvn7FqYFcgUg " =1203x664")

What do those folks do for a living? Those are two different websites even. So they flag `www.izzysoft.de` now, but not `android.izzysoft.de`? But the pattern matches again: An APK file, outside of our own PlayStore?!? And yes, it's the same file from 2 months ago. Which isn't even there anymore, since then. So, how often do they update their databases? A warning two months late, for something that's not even there, will certainly make you all feel very safe now, I guess. And scroll up to see how long it took them to find it. Hint: the file was originally added 2017-04-05 – 9 years ago. So you probably no longer even have the computer it “infected”…

**Update 2026-03-16:** of course, the review failed. Because they found out we have more APK files, so they just brought up the next. To continue, until our repo is empty? As before, *their own scanner* declared that APK as having no findings. Still, they libel not only our repo site, but the entire domain as “malicious” now.

**Update 2026-03-18:** still no reply to the reevaluation request. As that's all automatic, what takes them **that** long? Meanwhile I hear "PlayProtect" kicked in, and [*actively deletes these apps* from devices](https://github.com/sunilpaulmathew/AppVaultX/issues/6#issuecomment-4079105010), without asking for confirmation. Apps their own scanner declares clean. No proof given for why it is considered “malicious”. Transparency is clearly missing there, too. But whom to hold accountable? Where to intervene, how to obtain background? How should developers adjust their apps – or even know why they are considered “malicious”? I'd expect if the reason for such flagging is to get things fixed, those details would be made available. So is that really the reason?

**Update 2026-03-20:** finally a reply – the usual one: “Review failed”. Because of… what?

 !["Detected Issues" states "Harmful downloads", but still cannot given even a single example: "These URLs" is an empty list.](https://docs.katastima.org/api/files.get?sig=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJrZXkiOiJ1cGxvYWRzL2E2ZTRmNWNhLWJjNTktNDNkZS1iZGJhLTdhNjljNWU1Y2NlZi9iMWVkZWM0OC0wNTU5LTQ5MDQtOGM4ZS0yY2M4NmUzODU2NTYvaW1hZ2UucG5nIiwidHlwZSI6ImF0dGFjaG1lbnQiLCJpYXQiOjE3OTUyMzgxMDAsImV4cCI6MTc5NTMyNDUwMH0.bhkIilXA3Us1kev3y9FkNcM4sqKvmc_n71zhMEBruT4 " =928x243")

Go, fix the non-existing issue? Are we there again? 🤦‍♂

**Update 2026-03-25:** 4 days now since filing the last review request. Still no response. Filed another one.

## Let's do the same thing again, because why not?

**2026-04-14:** I guess they thought our monthly fun was due again – so they sent us this again:

 ![Familiar looking problem shown in Google's console](https://docs.katastima.org/api/files.get?sig=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJrZXkiOiJ1cGxvYWRzL2E2ZTRmNWNhLWJjNTktNDNkZS1iZGJhLTdhNjljNWU1Y2NlZi9kODY2YjdlNS01NTM3LTQxNTAtYmMyMC00OGJmOThkYWRkYmMvaW1hZ2UucG5nIiwidHlwZSI6ImF0dGFjaG1lbnQiLCJpYXQiOjE3OTUyMzgxMDAsImV4cCI6MTc5NTMyNDUwMH0.uWqc6pIv4fSMGqrH6GA6AIsu7osAIYE5RBNBWvOKt0Y " =930x452")

Sure, must be some invisible stuff I guess. And as someone finds it funny, they wanted to ensure “double fun”, sending the very same thing for another site as well. I've stopped asking myself what they do for a profession – and just requested a review, outlining that “all indicated pages have been cleaned”. True, because none were indicated. I'm so tired of this game…

**Update 2026-04-15:** an unexpected thing for a change. This time, we've been cleared within 24h. Is there hope to get that window to zero (with the false alerts closing themselves before they're opened)?

**Update 2026-04-19:** Same thing is back again. That was fast, too. And again, the list of pages is empty. So, reporting again I've cleaned them all…

**Update 2026-05-10:** Almost a month the truce lasted. Now they report the same file (uh? or not? cannot tell) again. Same ~~procedure~~ screenshot as the last one, so I do not repeat that here. And this “report” cleared **2026-05-11** after another “cleared all indicated links”.

**Update 2026-05-20:** Back again. I do not include a fresh screenshot, as it would be exactly the same (as was my review request, of course)… Same approach as last time, solved 2026-05-21.

**Update 2026-05-24:** They seem to be caught in a loop. Just 3 days, and the game starts again: Malware, but they don't know where. So I reported I cleaned – though I don't know where, either… And it looks like we're getting closer to self-destructing alerts: this one was cleared the same day even. But anyone placing bets how long until the next one?

**Update 2026-05-27:** The answer is: less than 3 days. Groundhog day. Same procedure as *every* year, James…

**Update 2026-05-31:** same game starts again: alleged malware, no URL. This time, "review failed" – alert is gone nonetheless. Make sense of that. Maybe we should simply ignore it, next time?

**Update 2026-06-12:** next round. Again, alleged malware, no URL.

**Update 2026-06-17:** not sure when they cleared the last one, but here comes the next. Or the same? Can't tell, not enough data…

**Update 2026-06-18:** yesterday's report cleared – or rather, replaced by another one. And once more, they cannot even tell websites apart:

 ![Once more Google reports on one website, what it found on another.](https://docs.katastima.org/api/files.get?sig=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJrZXkiOiJ1cGxvYWRzL2E2ZTRmNWNhLWJjNTktNDNkZS1iZGJhLTdhNjljNWU1Y2NlZi9mOTMxMjA0Ni00ZWU5LTRjMjMtOTUxNy0xYTZkOTBkMGVjYWUvaW1hZ2UucG5nIiwidHlwZSI6ImF0dGFjaG1lbnQiLCJpYXQiOjE3OTUyMzgxMDAsImV4cCI6MTc5NTMyNDUwMH0.qFfdX2F6pZmajEDo9p5JYlylVwm8lhCdcxmgmavuTwI " =1084x605")

(note that the reported file marks a false positive – known to be that for years, and the reporting malware engines obviously unable (or unwilling?) to fix their signatures – [see here](https://github.com/nfcgate/nfcgate/issues/176)) And before you ask: yes, they flagged **both** websites. Oh, and surprise (no sarcasm!): cleared within hours, wow!

**Update 2026-06-25:** next unknown malware. Review requested, comment just “same as last time”. Cleared 2026-06-26.

**Update 2026-06-29:** same again. Cleared on 2026-06-30.

**Update 2026-07-02:** same again. Cleared on 2026-07-03.

**Update 2026-07-06:** same again. Cleared the same day – with the message “review failed” :clown_face:

**Update 2026-07-16:** next one. Cleared 2026-07-17 as “review failed”. Maybe they meant “Review failed to find anything”?

**Update 2026-08-11:** wow, almost a month without a fake report. OK, got the next one today… (cleared 2026-08-12)

**Update 2026-08-26:** welcome back, invisible malware… (cleared 2026-08-27)  
**Update 2026-09-03:** next one – we must have missed the notification. Must have cleared itself somehow :man_shrugging:

**Update 2026-09-19:** another one. As it again says N/A for affected files, I’ve requested review with n/a as actions taken :zany_face: Should they one day come with a real security issue, I’m no longer sure I’d not just follow routine and ignore that as well… Cleared 2026-09-20

**Update 2026-09-27:** same again. Review requested with “n/a” again. Cleared 2026-09-28

**Update 2026-10-08:** same again. Cleared as “review failed” on 2026-10-09

**Update 2026-10-19:** same again. Cleared 2026-10-20

**Update 2026-11-02:** next one. Resolved as “failed” on 2026-11-03

**Update 2026-11-13:** Yay, next. They love us! Review failed successfully on 2026-11-14. Only to…

**Update 2026-11-17:** Repeat.

# Notes on how to disable Google Safebrowsing in browsers on Android

* For **Firefox**, see [How to disable ‘Deceptive content and dangerous software protection’ function for Firefox on Android](https://discourse.mozilla.org/t/how-to-disable-deceptive-content-and-dangerous-software-protection-function-on-android/96959/9)

  
  1. Access `about:config` from `chrome://geckoview/content/config.xhtml`
  2. Optionally, you can enable `about:config` permanently by setting `general.aboutConfig.enable` to `true`.
  3. Search for `safebrowsing` and toggle the option that may be relevant to the blocking. This requires some trial and error because there are so many possible reasons for the block. For my specific case, setting `browser.safebrowsing.malware.enabled` to false solves the problem.
* in **Brave** its in "privacy and security" setting
* **Vivaldi** on Android [does not use Safe Browsing (because usage there is restricted). Therefore there is no option to disable it](https://social.vivaldi.net/@Vivaldi/115967018521949789).
* **LibreWolf** has Safe Browsing disabled, due to concerns about censorship (matches our experience here)


# Related links

* they can even get you into a catch-22, as with some TLDs, the registrar might set a `serverHold` on your domain once Google puts you on the black list: [Never Buy A .online Domain](https://www.0xsid.com/blog/online-tld-is-pain)
* “Google Safe Browsing missed 84% of confirmed phishing sites in our dataset. […] Google is, quite literally, hosting phishing attacks on its own infrastructure and not catching them with its own detection tools”: [Huginn Report: February 2026](https://www.norn-labs.com/blog/huginn-report-feb-2026)
* [Google flags Immich sites as dangerous](https://immich.app/blog/google-flags-immich-as-dangerous) (10/2025) Immich is an open-source alternative to Google Photos. Quote: “… a single flagged subdomain would apparently invalidate the *entire domain*.” – exactly what happened here. At the bottom of their blog post, they list a bunch of additional cases: [Jellyfin](https://github.com/jellyfin/jellyfin-web/issues/4076), [YunoHost](https://forum.yunohost.org/t/google-flags-my-sites-as-dangerous-deceptive-site-ahead/20361), [n8n](https://community.n8n.io/t/deceptive-site-ahead-urgent/24152), [NextCloud](https://www.reddit.com/r/NextCloud/comments/w3x0fs/google_marked_my_nextcloud_app_as_a_dangerous/). Make your own connection; looks like each of them provides a service Google wants to have the monopoly on.
* [Google's Safe Browsing Protection in Chrome Goes Real-Time](https://tech.slashdot.org/story/24/03/14/1624218/googles-safe-browsing-protection-in-chrome-goes-real-time) (3/2024; they check live against a server-side list) Wish that (near) real-time was the case for their reviews…
* [RubyGems](https://github.com/rapid7/metasploit-payloads/issues/650) (2023) Comparable to our case: the same content the websites are blocked for is available at e.g. Github, but Github is of course *not* flagged. They come to the conclusion which we *suspect* as well: that this is "an intentional decision on Google's part". Further, also reflecting our experience: “Confusingly, VirusTotal has an entry for Google SafeBrowsing, and it claims that Safe Browsing finds this URL safe, while the Search Console claims that the URL is harmful.”
* [Google Safebrowsing has no way to appeal](https://news.ycombinator.com/item?id=40355635) (5/2024). It runs fully automated, no humans involved. Tracks. "Security review in Google Console does not get read by human, it just schedules automated website rescan - doesn't even say which non-hosted links are 'malicious' as a result. Feedback does not get read in a week too." And: “[Folks at Google] say there seems no internal appeal form or contact email for team which can actually fix this?”
* The [FAU.DE](http://FAU.DE) (Friedrich-Alexander-University) mirror of the [F-Droid.org](http://F-Droid.org) repository meanwhile (seen 2026-07-30) shows the following text if you try to download an APK file using a web browser:  
  »Unfortunately, Google abuses "SafeSearch" to flag our fdroid-mirror as "malicious" - because nobody is permitted to offer .apk-files for download other than the Google Play Store! As most browsers are dumb enough to trust "SafeSearch" to flag actually harmful content and not just competition that Google does not like, this makes our site completely unreachable in most browsers. As a result, we have had no choice but to censor away all the .apk-files. You can still download these files from within f-droid - but no longer with a browser.«

---

**Documents**

- [Google "Safe" Browsing](https://docs.katastima.org/s/iod-alleged-malware-reports/doc/google-safe-browsing-QzRVPEixQJ)
- [Netcraft](https://docs.katastima.org/s/iod-alleged-malware-reports/doc/netcraft-ZA6tnClxtM)