touzovitch

You have a point. Or even malicious links!

We have to be careful with the decrypted output. Redakt is an open source and collaborative project, just saying… 😜

Slow them down and prevent them to scale is actually not that bad. We are in the context of public content accessible to anyone, so by definition it can not be bulletproof.

Online Privacy becomes less binary (public vs private) when the internet contains content encrypted using various encryption methods, making it challenging to collect data efficiently and at scale.

Thank you so much for your comment though <3

I don’t think AI is bad as a whole. At least I would like to choose if the content I post online can be used (or not) to train models.

You are absolutely right! Using a single public encryption key can not be considered as secured. But it is still more than having your content in clear.

I intend to add more encryption options (sharable custom key, PGP), that way users can choose the level of encryption they want for their public content. Of course, the next versions will still be able to decrypt legacy encrypted content.

In a way, it makes online Privacy less binary:

Instead of having an Internet where we choose to have our content either “public” (in clear) or “private” (E2E encrypted), we have an Internet full of content encrypted with heterogeneous methods of encryption (single key, custom key, key pairs). It would be impossible to scale data collection at this rate!

Captcha was just an example :-)

What I’m trying to say is that any small changes that we add to the extension will have very few (or none) effect on the real users, but will force the srappers to adapt. That might require important human and machine ressources to collect data at a massive scale.

EDIT: And thank you for your feedback <3

r3d4kt-U2FsdGVkX1/lGJZ5fHhIJPQ8w7fdKIrvJKGa4C6hVzgxa99BNXMr7LQFL9Rur05EFVITe2pREZaianyq1F5k4dQEovbUKXWwjoj7R2ZXmu3z836vItVgTHh/Wen4p0pp&&&