Introduction:
In the digital era, technologies such as Artificial Intelligence — particularly Generative AI — are transforming how societies share and consume information. While these tools offer great potential, they also enable destructive practices like the large-scale spread of fake news and disinformation. Misinformation campaigns have evolved into coordinated efforts involving both state and non-state actors. Within this context, Foreign Information Manipulation and Interference (FIMI) has emerged, combining disinformation strategies with Artificial Technology (AI) to influence public opinion and undermine democracy.
This paper explores the intersection between Generative AI, Fake news, and FIMI, beginning with a conceptual definition of the terms. It then examines how AI technologies are used to create and disseminate disinformation. A case study on Russian FIMI during the war in Ukraine illustrates these dynamics in practice. The paper concludes by reflecting on AI’s broader implications for democratic information ecosystems.
By addressing those aspects, the paper aims to highlight the challenges of AI-based disinformation.
1. Generative AI and Fake News
This chapter introduces key concepts related to AI-driven disinformation, including definitions of Generative AI, Fake News, and Foreign Information Manipulation and Interference (FIMI). It further explores the implications of generative technologies for information dissemination, laying the theoretical foundation for understanding how AI can both enhance and undermine the integrity of public information.
1.1 Generative AI
In today’s fast-evolving technological environment, it is crucial to define Generative Artificial Intelligence (AI). Generative AI has emerged as one of the most transformative digital technologies of recent years, radically altering how information is created, circulated, and perceived. Its growing influence is not only felt in creative industries and everyday communication, but increasingly in the political and informational domains. In an era where public trust in information is fragile, understanding the mechanisms behind AI-generated content is crucial.
Unlike other forms of AI, which has primarily focused on classification or prediction, generative AI systems are designed to actively create content that resembles human-produced material (Ivanova & European Commission, 2021). These systems are often built on deep learning techniques, particularly transformer-based architectures, and are trained on large-scale datasets. As a result, generative AI can produce coherent text (e.g., through Large Language Models like ChatGPT), generate images from textual input (for example using diffusion models such as DALLE), and create synthetic audio, including voice and music. The ease with which believable yet entirely fabricated content can be produced undermines the distinction between real and synthetic, thereby facilitating the spread of disinformation, manipulated media and contributing to the erosion of public trust in digital information environments.
While generative AI provides new means of producing content, its societal impact is shaped by how these tools are employed in the construction and circulation of false or misleading narratives. In the next section, the concepts of Fake News and Foreign Information Manipulation and Interference (FIMI) are explored in order to understand how technological capabilities are transformed into instruments of influence and manipulation.