Romanian Cyber Security Journal / Spring 2024, No. 1, Vol. 6

Between the Lines: Generating, Detecting and Defending against Textual Deepfakes

Elena-Anca PARASCHIV, Carmen-Elena CÎRNU

---

Abstract

This paper provides an in-depth exploration of textual deepfakes, emphasizing their significance in the realm of cybersecurity. It delves into the sophisticated techniques employed by artificial intelligence (AI) algorithms to enerate highly deceptive textual content, highlighting the potential cybersecurity threats they pose. Through the examination of malicious actors’ utilization of textual deepfakes for activities such as misinformation campaigns and phishing attacks, this paper highlights the critical need for robust detection mechanisms and defense strategies. By fostering awareness towards the impact of deepfakes, enhancing deepfake detection capabilities, and advocating the responsible usage of AI technologies, this paper aims to address the cybersecurity challenges posed by textual deepfakes and safeguard the integrity of digital communication channels.

Keywords

GPT, AI, Deepfakes, Textual Deepfakes, Cybersecurity