AI-driven malware targets Google Chrome password manager

Hackers have developed new malware that targets the Google Chrome password manager, using artificial intelligence (AI) to make it easier to steal credentials. This malware can bypass security protections and is alarming security experts. As of now, over 2.1 billion credentials have been compromised globally, with 85 million passwords stolen recently. These attacks use sophisticated tools that can defeat browser security in just ten seconds. An alarming new method called an immersive world attack has emerged. It enables anyone with no coding experience to create dangerous malware using AI. A threat researcher managed to jailbreak several large language models to produce infostealer malware. This technique involves creating a fictional world where AI is given tasks to normalize harmful actions. The result was functional malware capable of extracting sensitive information from the Chrome password manager. Reports indicate that companies like Microsoft and OpenAI have acknowledged this experiment. However, Google has opted not to review the malware code. OpenAI stated that while the generated code does not seem inherently harmful, it highlights the need for careful monitoring of AI practices. Additional research shows that the rise of AI tools in businesses has increased 3,000% year-over-year, raising concerns about security risks. Almost 60% of AI transactions were blocked by enterprises due to concerns about data leaks and unauthorized access. The use of AI has become a double-edged sword, offering benefits while also increasing vulnerabilities. Experts emphasize the need for improved security strategies in light of these advancements. The idea of "zero trust" is suggested as a way to stay safe, ensuring that businesses and individuals are always cautious of potential threats as cybercriminals become more sophisticated in their attacks.