In our digital age, the news is often accused of bias. But can artificial intelligence help us understand the political leanings of different newspapers? A recent study put several large language models (LLMs) to the test, asking them to rate the political orientation of articles from 40 newspapers worldwide. The results were surprising. The LLMs, including popular names like ChatGPT and Gemini, showed significant disagreement in their assessments. Some models leaned heavily towards labeling papers as left-leaning, while others favored a right-wing classification. This inconsistency raises important questions about the reliability of using AI to detect bias. While some LLMs clustered newspapers towards the center, others showed wide variations, often contradicting established understandings of the publications' political stances. The study highlights the challenges of using AI for nuanced tasks like political analysis. The researchers emphasize the need for further investigation and development of LLMs to reduce biases and improve accuracy. They also call for human experts to help benchmark and validate the models' assessments, creating a collaborative effort to improve AI's ability to understand the complexities of political discourse. While the study reveals current limitations, it also points to the potential of AI to play a positive role in journalism. By improving media literacy, supporting quality journalism, and assisting with fact-checking, LLMs could help combat misinformation and foster a more informed public discourse. The future of AI in journalism depends on addressing these challenges and working towards more reliable and unbiased models. This research is a crucial first step in understanding how AI can help us navigate the increasingly complex world of news and information.