A recent study by the security company Irregular warns that passwords generated by major AI models like ChatGPT, Claude, and Gemini are highly predictable and should not be used. According to Sky News, the founder of Irregular, Dan Lahav, stressed the importance of changing any AI-generated passwords immediately, as the issue is not widely recognized among users. Predictable passwords can be easily guessed by automated tools that hackers employ, posing significant risks to personal and corporate security.

The research highlights a substantial difference between AI-generated passwords and those created by dedicated password management services. While services like Apple and Google’s password generators use secure methods to produce truly random and complex passwords, AI models rely on their training data, leading to less secure outputs. The predictability of these passwords empowers cybercriminals, making them an undesirable choice for safeguarding sensitive information.

Lahav's statements draw attention to the broader implications of using AI in cybersecurity. As artificial intelligence continues to permeate various aspects of daily life, the risks associated with its misuse extend beyond just password generation. This awareness emphasizes the need for users to be vigilant and knowledgeable about the tools they employ for online security, ensuring they do not unknowingly compromise their personal data through reliance on inadequate security measures.