←back to thread

Comet AI browser can get prompt injected from any site, drain your bank account

(twitter.com)
645 points helloplanets | 1 comments | 24 Aug 25 15:14 UTC | HN request time: 0.226s | source
Show context
charcircuit ◴[24 Aug 25 15:52 UTC] No.45005190[source]▶
Why did summarizing a web page need access to so many browser functions? How does scanning the user's emails without confirmation result in being able to provide a better summary? It seems way to risky to do.

Edit: From the blog post for possible regulations.

>The browser should distinguish between user instructions and website content

>The model should check user-alignment for tasks

These will never work. It's embarrassing that these are even included, considering how models are always instantly jailbroken the moment people get access to them.

replies(5): >>45005229 #>>45005325 #>>45005382 #>>45005446 #>>45006024 #
1. esafak ◴[24 Aug 25 16:13 UTC] No.45005382[source]▶
Beside the security issue mentioned in a sibling post, we're dealing with tools that have no measure of their token efficiency. AI tools today (browsers, agents, etc.) are all about being able to solve the problem, with short thrift paid to their efficiency. This needs to change.