A new vulnerability in browsers that use artificial intelligence, such as Comet from Perplexity, makes hackers hack the browser through Link! Let me explain to you then what happened?

👈The information security team at LayerX found browsers like Comet from Perplexity that use artificial intelligence in the browser to help the user read the content and communicate with your email. They found a loophole that hackers who place a Prompt Injection in the URL or website address, the browser reads this speech, translates it, and executes it on itself!

I mean, let's say that the hackers sent you or you found an article with a link that you clicked on. What happens is that the browser reads this link and puts in it a parameter that says the prompt it wants to execute. What happens after this is that the browser reads this prompt through the AI Model that the browser uses and executes it on its own browser. Then the attacker can access e-mail messages, view the files attached to the e-mail, and steal anything within the browser's scope, such as Cookies/Sessions, etc., but He did not hack your entire device (he only hacked the browser), or let's say he uses the model in the browser to send him these things stored in the browser itself! This is a big problem.

#linux_network_tools #network_tools #chatgpt #network_monitoring #cybersecurity #AI #programmer_idea #data_protection #cloud_backup #data_recovery #comet