Interesting and smart approach - most noise generators are obviously artificial in their traffic patterns.
I've been thinking about browser privacy from a different angle: not hiding what you browse, but hiding what's visible on your screen when you share it. Screen sharing during video calls basically bypasses every privacy tool you have running (VPN, tracker blockers, etc.) because the other person sees your raw screen.
The layered privacy defense framing makes sense. This handles the ISP/tracking side. But what handles the "accidentally showed my email to my entire team during a screen share" side? Different threat model but equally common.
Hey, smart one!
- Why not allow the behavioral/timing intervals somehow random, with predefined threshholds? Just to simulate real user interactions.
- Also, what’s the criteria for choosing those URLs? (Any security measures?)
- Those URLs being visited are the same? i.e. you mentioned the URLs being visited are hardcoded, so eventually the ISP can filter our those URLs… unless if they are continuously enriched/interchanged.
It does follow links as to randomize it some but yes the hardcoded list or urls is a limitation and possible pattern. This list can be changed and added to (accepting PR's).
Interesting and smart approach - most noise generators are obviously artificial in their traffic patterns.
I've been thinking about browser privacy from a different angle: not hiding what you browse, but hiding what's visible on your screen when you share it. Screen sharing during video calls basically bypasses every privacy tool you have running (VPN, tracker blockers, etc.) because the other person sees your raw screen.
The layered privacy defense framing makes sense. This handles the ISP/tracking side. But what handles the "accidentally showed my email to my entire team during a screen share" side? Different threat model but equally common.
Congrats on shipping.
Hey, smart one! - Why not allow the behavioral/timing intervals somehow random, with predefined threshholds? Just to simulate real user interactions. - Also, what’s the criteria for choosing those URLs? (Any security measures?) - Those URLs being visited are the same? i.e. you mentioned the URLs being visited are hardcoded, so eventually the ISP can filter our those URLs… unless if they are continuously enriched/interchanged.
It does follow links as to randomize it some but yes the hardcoded list or urls is a limitation and possible pattern. This list can be changed and added to (accepting PR's).
I wonder if it will ”flag” you in cloudflare and google so you will end up in captcha hell.
I haven't seen that so far!