Browser Extension Integration for StopSlop

TomatootamoT

Browser Extension Integration
Develop a browser extension that highlights AI-generated content across the web, not just within Kagi Search. This would extend SlopStop's utility beyond search results and help users identify potentially deceptive content while browsing.

You can see my full wish list here:
https://kagifeedback.org/d/8921-stopslop-additions

tux0r

Instead, I believe an API would suffice, which would reduce the effort for the Kagi team and be even more practical for the community.

For example:

Someone develops a browser extension and/or a userscript that does exactly what you are suggesting here.
Someone develops an Emacs package that can be integrated into, for example, mu4e, making it easy to detect and filter out AI-generated emails.
This would then also be possible for Usenet, IRC, etc.

TomatootamoT

While that would be good for the internet i also think having it be under the Kagi name would be greatly beneficial for gaining positive press. But i think if kagi decided to open StopSlop's Api it would include the extension code ect. Both could be possible but i think that's up to the team. Thanks for the idea.

tux0r

That could be solvable by the API's terms of use requiring that Kagi's name must be mentioned; or by the API being mandatorily linked to a user account, so that one has to be a Kagi user to use the various packages – that would automatically be (very good) advertising for Kagi. 🙂

Thibaultmol

{why not both spanish meme}

They're still looking into how exactly they'll make the data available but think it makes sense since it's Kagi initiative that Kagi also has an extension for it or some other thing. But an API will probably be considered as wel

Perdreau

In France, the Next.ink website made a great analysis of AI slop news site and built a chrome and firefox extension to detect this website and warn users.
in fact there is now more than 8500 websites that are detected as AI Slop in French just one year after their launch.
https://next.ink/208687/extension-next-des-alertes-sur-8-500-sites-genai-et-contre-les-attaques-par-homographes/
https://next.ink/195816/alerte-sur-les-sites-genai-notre-extension-signale-plus-de-6-000-sites-et-passe-en-v2-3/

The most difficult job for them is that they obscure the website lists with bloom filter because in one year they see some inventive people change domain names to bypass detection.
The source of problem are multiple :

Adsense websites that use SEO and expired domain names to be in the top of Google Discover for some keywords
Stranger ingerence website that use fake news to destabilize country with fake news (they are well documented here :
https://github.com/VIGINUM-FR/Rapports-Techniques
https://www.rts.ch/info/monde/2025/article/faux-sites-d-info-locale-la-propagande-pro-russe-s-infiltre-en-france-29040599.html
https://www.irsem.fr/storage/file_manager_files/2025/10/focus-3-charon-a4-ok.pdf
financial scam or other fake investments or medical threats

I tried to import some of them before ai slop launching in my personal block/downgrade list but because of the limitations of 1000 domain names it's not easy...