Many people are excited to transition to the new era of the internet thanks to AI web browsers like ChatGPT Atlas, Comet, and even Google Chrome. Users have good reason to rethink their decision or even exercise caution when utilizing these AI browsers, nevertheless. Security experts are concerned that if there is a high demand for these browsers, hackers may try to get around their problems, which might make it simple to steal important data.
Millions of people find AI browsers to be a tempting alternative due to their abundance of capabilities and ease of use, which many claim allows for multitasking at an entirely new level. However, that convenience might turn out to be expensive, and experts have expressed concern about this.
Although cool, AI browsers are not secure.
Prompt injections are the primary threat to AI browsers such as ChatGPT Atlas and Comet. We are all aware that AI chatbots react to commands, whether they are to create an image, place an order for groceries, or even ask the AI to look through a paper.
For example, the payment information is probably included when you use AI to shop for groceries, and these malicious actors may utilize prompt injections to force the AI model to display those figures. It’s that easy: these injections can also target users who only instruct the AI to summarize a malicious webpage.
With the rise of AI browsers, these assaults are unavoidable, and hackers will be salivating over the potential of AI and how it can be exploited by inserting prompts into web pages to trick the AI model and cause it to perform tasks incorrectly. The security flaws in Perplexity’s Comet have already been examined, leaving the possibility of rapid injections open.
Artificial intelligence has been known to experience hallucinations and provide information that isn’t even available online.
Because of these incidents, there is a high likelihood that AI browsers may go rogue. Brave and other platforms that investigate these browsers have alerted users to the dangers of these programs. The business stated in a statement that “the risks multiply as users become accustomed to AI browsers and start trusting them with sensitive data in logged in sessions—such as banking, healthcare, and other critical websites.”
They Require Your Information
Have you ever considered how these artificial intelligence browsers might manage your work and automate some of your routines? To make it work, they require a great deal of access to your data. Not simply the standard contact list, phone number, and email address, but much more complex. This covers reading your emails, granting third-party task plugin access, viewing your browsing history, tracking your online activities, and more.
Be Astute and Wary
AI browsers are clearly dangerous, and certain proof-of-concept demonstrations highlight how serious their effects are. However, that does not imply that people should be wholly opposed to utilizing new technologies.
Every generation has witnessed the leap, and there are risks in every area. The only things people need to be mindful of are how they use these AI browsers, how much information they give the AI model, and which jobs they truly want automated.
