Bots with Polite Manners Fool Most People on Social Media

Researchers at Surfshark analyzed 1,722 participants worldwide, testing their ability to separate human comments from AI-generated ones in a social media setting.

Bottom line: Polite AI-generated comments are more effective at deceiving humans on social media than those that are rude or aggressive.

What's happening: Surfshark's study found that 55% of participants were unable to distinguish between human and AI comments, with the majority being fooled by comments with a polite tone. The company analyzed comments generated by AI using Natural Language Processing (NLP) techniques. The participants were exposed to 1,722 comments, including 112 AI-generated comments with polite language.

What to do: Security leaders should prioritize testing their social media monitoring tools for AI-generated comments, particularly those with polite language, to prevent potential phishing attacks. They should also consider implementing a human-in-the-loop review process to verify the authenticity of comments.

Source: Help Net Security