A prompt-injection vulnerability in the AI assistant allows attackers to create messages that appear to be legitimate Google Security alerts but instead can be used to target users across various Google products with vishing and phishing.
Related Posts
What are the shortest words with a given number of…