@waystar – Happy to try and help.
I am not quite sure I fully understand the question so I want more info to be sure I answer thouroughly. That said currently our restrictions can be applied based on not only specific user roles but the pro version also can apply them based on ecommerce purchase history. That can’t be faked with any “bypass”.
The content is simply not rendered to the browser if the user doesn’t have the correct access. So even a scraper will see nothing.
This applies to the Rest API as well. So assuming your registration & purchase forms are secure there shouldn’t be a way to bypass it. That would require a backdoor or leaking of the content in some way.
If there are things we can do to prevent AI scraping entirely we are happy to look at those also for future updates.
Lastly if a person has access to your content and then lets LEO scrape the page while viewing it, I’m not sure there is anything that can be done to stop it, they are using a browser extension with raw access to the DOM elements. The only way it could skip some content and not others is if the developers made a way to designate content to be ignored, otherwise it just reads everything.
Also from their own FAQ:
“With respect to the servers that power Brave Leo, your browser will share your latest prompt, the context of your current conversation, and—when the use case requires it—the necessary context from the page you’re viewing. Note that once a chat is closed, all records of that chat are erased We do not collect identifiers that can be linked to you (such as IP Address), and no personal data is retained by Brave-hosted AI models. Brave Leo offers several models to choose from. If you select a model from Anthropic and submit Leo queries, that data will be processed by Anthropic for a period of 30 days, then deleted.”
~Lastly if a person has access to your content and then lets LEO scrape the page while viewing it, I’m not sure there is anything that can be done to stop it, they are using a browser extension with raw access to the DOM elements. The only way it could skip some content and not others is if the developers made a way to designate content to be ignored, otherwise it just reads everything.~
That’s exactly the problem. ILLEGAL COPYRIGHT INFRINGEMENT or worse IP THEFT. Allow me to explain from our point of view. 100K project management hours under belt. Most AI models are coping the entire internet. but what are they really copying?
They are Copying a Copy of a Copy of a Copy of a Copy of a copy >> infinity. In the end what results will spit out? INTELLIGENCE?
It will spit our the “MEAN,’ bias general opinions nothing close to anything called intelligence.
And this is why most AI models are now failing. bc they need data that has not been exposed to the internet in order to STEAL. aka rampant IP theft.
Brave leo scrapes a page and gives a summary. This is the beginning of Spinning text. And if third party browsers are allowed to scape this is a ILLEGAL BLATANT BREACH of our right to privacy and intellectual property theft.
you see these thieve need Valuable content else their data (aka via data brokers) is useless.
now that I stated the facts LETS STOP IT IN ITS TRACKS and we can help if needed.. Let us know.
BTW ty for the well thought out response.
these are extensions to browsers that are illegal to actually use.
So now we must block entire browsers?? Hmmm..
Btw use case is fantasy case use is reality and worlds apart
Waystar
hello presale question.
For pro or commercial product support please contact the developer directly on their site. This includes any pre-sales topics as well.
As the developer is aware, commercial products are not supported in these forums. I am sure they will have no problem supporting you there.
@danieliser Can you provide a contact URL for this user? This is not a conversation for here.
TY for your time, response and help! Time to code a solution:)
@moderator – This isn’t really a question about the product at all, not a pro or free question. Seems more hypothetical what can we do type of thing. Our pro product doesn’t offer anything to this users question if that makes since. I’m just trying to answer in the most helpful way.
@waystar – In reality the philosophical question is not one easily solved, nor one we can provide a technical solution for via our plugin.
The truth is, if they can view the content, there is no stopping them from feeding it into the AI generator of their choice once they copy/paste it or even screenshot and OCR the text.
What your asking for would actually require the AI scraping tools themselves to respect some pre-defined boundary. Then we could implement that boundary on the protected content.
As for unauthed users seeing the content, that shouldn’t occur with our plugin, it doesn’t leak the protected content to the page hidden somewhere if that makes since.
If you find some W3C Standard for marking up HTML in a way that AI’s respect, we can definitely add it.