Authors

Description

The talk addresses the presentation of our recent third book on artificial intelligence and cybersecurity “Cybersecurity: data has the answer.” The proposal seeks to introduce the importance of cybersecurity and how to develop artificial intelligence models through six projects which seek to guide the viewer how to use machine learning from both a defensive and offensive perspective. The projects deal with the results of generative text models (LLM) for cybersecurity and prompt hacking, an AI that starts from the behavior of ransomware to deceive another AI, an AI Pair programmer system for detecting vulnerabilities in real time, an AI to orchestrate OSINT tools, and a practical case of fraud detection in a company in Colombia. In addition, the presentation will provide an introduction to the area through a ten-year research history, in which some of the results achieved in the first two books published will be shown ("Cybersecurity: an approach based on data science", "Cybersecurity: data has the answer"). During the presentation, the key concepts will be given, a brief introduction to the three books, a series of demos of the chapters will be presented (for example, AI for OSINT, Prompt hacking to ChatGPT, audio deepfake) and finally a demo about the current work we have been doing on a data acquisition system for social engineering using audio and text. The talk seeks to present the previous results and link how the application of Python has allowed their development.