California (Web Desk): A new study conducted in the United States has revealed that advanced artificial intelligence tools can analyse personal information available on social media and create fake messages that appear genuine and trustworthy.
According to the research, artificial intelligence models can use users’ interests, daily routines, travel experiences, photographs and social connections to generate personalised deceptive messages. For example, a person could receive a special offer from a hotel or information about a concert in which they had previously shown interest.
Researchers from the University of Texas at Arlington and Louisiana State University have termed this method “contextual phishing”. During the study, they examined the public accounts of 200 Instagram users and found that around 10 posts by a user could provide enough information to create personalised fraudulent messages.
The researchers used different artificial intelligence models to generate approximately 18,000 messages and evaluated them against eight different criteria, including their relevance to personal information, perceived authenticity, persuasiveness and emotional impact.
The findings showed that messages generated through artificial intelligence could appear more genuine and less suspicious than conventional phishing attacks. Some security systems also failed to identify them.
The researchers have warned that even publicly available information routinely shared on social media could be exploited by cybercriminals to carry out personalised scams and online attacks.