Team The News Miners from the Babes-Bolyai University in Romania, answers this using over 7,000 news articles from The Guardian covering 15 years of AI-related reporting (2009 - 2024).
They earned a spot among the top 15 teams worldwide in the 2025 global student competition, The Curiosity Cup.
They used:
-
OpenAI’s GPT model to detect whether articles framed AI as an opportunity, threat, or neutral
-
SAS Viya for Learners to analyze text, extract topics, and visualize trends.
What did they discover?
-
Security is often framed negatively, and as many as 78% of articles show AI as a threat
-
Healthcare has a more positive framing, 43% portray AI as an opportunity
-
Technology and finance show mixed or mostly negative sentiments
-
A huge rise in articles about Responsible AI: from <5% before 2022 to ~20% in 2023 to 2024
Trends over time:
-
Early years: Focus on automation, surveillance, and the economy
-
2014 – 2018: Rise of social media, consumer tech, and healthcare
-
After 2023: Surge in topics like LLMs (e.g. ChatGPT), ethics, and regulation
Conclusion: Media narratives play a crucial role in shaping society's perception of AI.
As AI becomes more powerful, ethical and balanced reporting is more crucial than ever.
Great job: Andrei Malan, Andreea Popoviciu, Dragota Iulia-Mirela, Alexandru Sorocan, and their Faculty Advisor, Darie Moldovan, for bringing clarity to such a timely topic!
To view all relevant graph visualizations and read the full project paper, see the PDF linked below the main blue article headline.
