Disinformation represents one of the greatest challenges of modern information society as it affects public opinion, political processes and overall trust in the media.
The aim of the thesis is to explore the impact of AI in recognizing and preventing disinformation. The focus is placed on the detection of deepfakes and the use of automated fact-checking systems. As part of the review of existing solutions I present the basic concepts of deep learning and the current state of deepfake detection. Furthermore I examine the functioning of fact-checking systems, their databases and the role of automated solutions in this field. Current solutions show promising results but face significant challenges such as poor generalization of detection models, high computational demands and the lack of high-quality datasets. AI can greatly contribute to the development of tools for detecting and preventing disinformation. However, raising public awareness and promoting media literacy will be equally important for ensuring long-term effectiveness.
|
