Measuring the Impact of Fear-Inducing Headlines

The study employed multiple sentiment analysis tools to assess the emotional tone of the headlines, including VADER (a lexicon and rule-based sentiment analysis tool that scores words from most extreme negative to most extreme positive), AFINN (a tool that assigns scores to words with negative scores indicating negative sentiment and positive scores indicating positive sentiment), and TextBlob (a library that provides a simple API that returns a polarity score indicating a negative sentiment, a positive sentiment, and neutrality).

The results showed a marked presence of negative sentiment, with approximately 40% of the headlines conveying fear or anxiety about AI. These findings suggest that the media is not just reporting on public fears about AI but is also contributing to their amplification.