I have checked a variety of languages, and it seems to happen consistently for all Latin-script language Wikipedias. Some examples:

• Norwegian Bokmål – https://stats.wikimedia.org/#/no.wikipedia.org/reading/total-page-views/normal|bar|2-year|agent~user|monthly
• Norwegian Nynorsk – https://stats.wikimedia.org/#/nn.wikipedia.org/reading/total-page-views/normal|bar|2-year|agent~user|monthly
• Icelandic – https://stats.wikimedia.org/#/is.wikipedia.org/reading/total-page-views/normal|bar|2-year|agent~user|monthly
• Faroese – https://stats.wikimedia.org/#/fo.wikipedia.org/reading/total-page-views/normal|bar|2-year|agent~user|monthly
• Swahili – https://stats.wikimedia.org/#/sw.wikipedia.org/reading/total-page-views/normal|bar|2-year|agent~user|monthly
• Cebuano – https://stats.wikimedia.org/#/ceb.wikipedia.org/reading/total-page-views/normal|bar|2-year|agent~user|monthly
• Waray-Waray – https://stats.wikimedia.org/#/war.wikipedia.org/reading/total-page-views/normal|bar|2-year|agent~user|monthly
• Aymara – https://stats.wikimedia.org/#/ay.wikipedia.org/reading/total-page-views/normal|bar|2-year|agent~user|monthly
• Inari Sámi – https://stats.wikimedia.org/#/smn.wikipedia.org/reading/total-page-views/normal|bar|2-year|agent~user|monthly
• Igbo – https://stats.wikimedia.org/#/ig.wikipedia.org/reading/total-page-views/normal|bar|2-year|agent~user|monthly (with an even bigger spike in September, unlike the others)

Compare to languages with non-Latin scripts:

• Yiddish – https://stats.wikimedia.org/#/yi.wikipedia.org/reading/total-page-views/normal|bar|2-year|agent~user|monthly
• Armenian – https://stats.wikimedia.org/#/hy.wikipedia.org/reading/total-page-views/normal|bar|2-year|agent~user|monthly
• Egyptian Arabic – https://stats.wikimedia.org/#/arz.wikipedia.org/reading/total-page-views/normal|bar|2-year|agent~user|monthly
• Standard Moroccan Tamazight – https://stats.wikimedia.org/#/zgh.wikipedia.org/reading/total-page-views/normal|bar|2-year|agent~user|monthly
• N'Ko – https://stats.wikimedia.org/#/nqo.wikipedia.org/reading/total-page-views/normal|bar|2-year|agent~user|monthly

(Curiously, the Dhivehi Wikipedia – https://stats.wikimedia.org/#/dv.wikipedia.org/reading/total-page-views/normal|bar|2-year|agent~user|monthly follows the Latin-script pattern.)

Maybe this is related to this problem that I reported a month ago?

Hi there! Any progress? I am from afwiki.

Desktop "viewers" from countries where few people speak the language – the Philippines, India, Japan:
https://stats.wikimedia.org/#/hr.wikipedia.org/reading/page-views-by-country/normal|map|last-month|(access)~desktop*mobile-app*mobile-web|monthly
https://stats.wikimedia.org/#/hr.wikipedia.org/reading/unique-devices/normal|line|1-year|access-site~mobile-site*desktop-site|daily

https://stats.wikimedia.org/#/sv.wikipedia.org/reading/page-views-by-country/normal|map|last-month|(access)~desktop*mobile-app*mobile-web|monthly

Hi @jhsoby-WMNO would you be able to assist us in getting this ticket assigned to someone? This is making our stats unreliable.

Hi Guys! Our January 2026 pageviews are even worse! We need your assistance! Regards!

Hi @Milimetric and @JAllemandou, would you be able to assist us in getting this ticket assigned to someone. The problem is causing wikistats to no longer be a reliable source of page views and related stats. We don't have an alternative source for these stats, and are not sure what else to do about it.

Your assistance would be greatly appreciated.

Ping @Ahoelzl and @GGoncalves-WMF for prioritization.

Thank you so much for pinging. This is suspicious indeed, and especially interesting is @jhsoby-WMNO's finding about non-latin wikis.

@Hghani points out (on a first look) that this can be related to T413027. That ticket also hints at some signals we can incorporate into our classification from here.

Hi there! User Oesjaar from the Afrikaans Wikipedia. Our pageviews for February is back to normal. Will the pageviews for Nov, Dec & Jan remain as it is? Regards

Hi, we're still investigating this in T416933 and haven't made changes to the underlying data, so the trend you're seeing for February is due to traffic changes rather than any updates in our classification.

We're deciding whether in this case we'll update our detection and re-classify the impacted data, or leave it as-is and put annotations on Wikistats to flag that an incident happened (but we won't leave it as-is without any annotation).

At the moment, we have put an annotation in the pageviews page to note that there is an ongoing investigation.

Sharing what I've heard from @GGoncalves-WMF and others working on this: we have devised a new bot detection rule that captures most or all of this spike. In addition to applying it to new data from the start of April, we have decided to retroactively apply it to December–March data: T421735 (we can't apply it to November as we no longer have the source data for that period).

Thank you for the feedback . Appreciated

Hi there! Our views for April was back to normal. Regards!

For the record: the data correction applying the new bot detection rule to December–March data has finished (T421735), and the effects can be seen in Wikistats. We will publish a data issue overview page on Wikitech soon.

I've updated the ticket to include the link to the data issue report:

ℹ️ Update 2026-06-15: the data issue report about the spike in traffic and our resolution has been posted to Wikitech: Nov 2025 spike in bot traffic