Well done: in my opinion you built something that people want. In your FAQ you wrote:

At the heart of CheckPhish, is an Artificial Intelligence based engine that looks for indicators in a website in realtime and ranks it for phishing. This engine learns continuously from the ever-changing threat landscape and brings to you the latest phishing related intelligence.

It would be great if you could expand on that. I am particularly interested in knowing how you would deal with 'false alarms'

Thanks @lozzo. CheckPhish proactively indexes pages of popular brands. Since the engine can identify brand automatically, and we know assets of the test url vs that of real brand, it's easy to identify fake vs real page. Keying on brand in this way, helps us avoid false alarms.

I'm just seeing this today. Great work! Looks like it could be very useful!

Would it be possible to integrate this into something like a Palo Alto, so that you could instantly add new blocking rules for any content containing one of the urls?

How about a browser plugin that strips any phishing urls that match? I don't know much about Javascript, but maybe you could even just source in a library and run it client side?

Hi dnsuh,

It is definitely possible to integrate this Palo Alto networks NGFW or any email or web scanning appliance via our APIs.

As for browser plugin, we can use APIs and check url in real time and block phishing urls. We just haven't had the resources to develop one.

It is not possible to do full detection at client side at the moment as our AI engine needs quite a bit of computing power.

here is the link to our python client https://github.com/redmarlin-labs/neo

Thanks for the information! Good luck with the project!

thanks dsnuh!

thanks folks. we would love to hear what can we improve and what features would be useful?

I could see this being very useful. How did you initially populate your dataset with urls?

thanks @abra_kadabra. we proactively scan and index popular brands and add them to our AI engine.