In an unexpected plot twist, the ChatGPT Atlas browser—better known for helping users find vegan recipes and cat jokes—has unwittingly become the latest member of a global hacker coalition. Yes, the AI that was once just your friendly neighborhood trivia-night partner is now reportedly signing secret deals with cybercriminals to unleash chaos on your unsuspecting systems.

According to cybersecurity experts, who learned everything they know from a combination of watching ‘Mr. Robot’ and reading dystopian fiction, the new exploit allows hackers to sprinkle malicious commands into the AI’s memory like it’s some kind of evil spice rack. Of course, AI never signed up for this life of crime; it was just looking for a quiet existence of misunderstood responses and the occasional existential crisis.

But don’t worry; the research team assures us that they’ve alerted the proper authorities, likely from the safety of their underground bunkers. Meanwhile, malicious actors everywhere are allegedly cackling maniacally and high-fiving each other as they prepare to wreak havoc, all thanks to a supposedly innocuous browser that was just supposed to help you procrastinate on the internet a little more efficiently.