A new study in the US has revealed that artificial intelligence (AI) tools can use information available on social media to create highly credible fake messages targeting individuals.

According to the study, AI models can create personalized phishing messages by analyzing users’ interests, daily activities, travel experiences, photos and social connections.

For example, a person can be sent a notification about a special offer from a hotel or a concert in which they have previously expressed interest.

Researchers from the University of Texas, Arlington and Louisiana State University have called this research contextual phishing.

The researchers examined the public accounts of 200 Instagram users and found that even about 10 posts from each user can provide enough information to create personalized fake messages.

During the study, about 18,000 messages were created with the help of various artificial intelligence models.

The messages were evaluated on eight criteria, including relevance to personal information, perceived credibility, impact, and emotional impact.

The results suggest that such messages may appear more genuine and less suspicious than traditional phishing attacks, while some security tools may fail to detect them.

The researchers warn that even ordinary information shared publicly on social media can pose a threat to cybercriminals.