AI-Enhanced Attacks

In November of last year, the latest version of OpenAI’s AI-based bot, “ChatGPT,” was released for general use. We can say that with the arrival of “ChatGPT” in our lives, a serious change has begun not only in the IT world but in every environment. We are still at the beginning of this… At our company…

In November of last year, the latest version of OpenAI’s AI-based bot “ChatGPT” was released for general use. We can say that with the entry of “ChatGPT” into our lives, a serious change has begun not only in the IT world but in every environment. We are still at the beginning of this topic… Our expert friends in our company have already started writing programs using “ChatGPT”. The first attempts yield surprisingly positive results. As someone extremely sensitive about the ethical use of technology, we are a team that utilizes “ARTIFICIAL INTELLIGENCE” in “ Cybersecurity ” applications. However, could the opposite of this utilization also be possible? While researching and working with the suspicion that attacks could be made using artificial intelligence, as soon as I read the article published by the esteemed Scientist Prof. Dr.-Ing. Sebastian Biedermann on security-insider.de on 11.01.2023 , I wanted to agree with his important statements and immediately translate his writings into Turkish and share them as they are. What is written should not worry us, but as I carefully express at every opportunity, technology needs to be used by integrating not only with “ Science and Reason ” but also with “ Conscience and Ethics ” values, and attention should be paid to this issue. I would like to emphasize that all of us, as technology producers, bear a serious responsibility in this field, especially in the 100th anniversary of the Republic of Türkiye. I thank Prof. Dr. Sebastian Biedermann very much and share his article in Turkish below. Sincerely, Cüneyt Kalpakoğlu IT security benefits from artificial intelligence, but is this always the case? Artificial intelligence “ChatGPT” has been on everyone’s lips since November. Initial tests show that dialogues with the impressive chatbot are suitable not only for entertainment but also for preparing sophisticated phishing campaigns. This could be a new problem in the field of IT security. Artificial intelligence is advancing ever faster and gradually conquering all areas of information technologies. Although the topic was initially scoffed at by many and did not receive the attention it deserved, most skeptics are now also impressed by the various results of modern AI systems. Some of the current AI technologies are also frightening, and the question arises as to what else might await us in the next few years. Artificial intelligence to improve IT security Artificial intelligence is also increasingly entering the field of IT security. With the help of artificial intelligence, modern attack detection systems learn to distinguish “normal” network traffic and “normal” events of a running operating system from anomalies and to intervene accordingly. Sandbox systems use artificial intelligence to automatically detect and classify still unknown potential malware with the help of dynamic analyses. Although there are many technical applications in this area that actually benefit from artificial intelligence and experience improvements as a result, it seems to me that the buzzword “artificial intelligence” is also often used for marketing purposes. Artificial intelligence itself as a target In addition to many new possible uses, the last few years have shown that AI systems themselves are not invulnerable. AI models can leak detailed information about your traini…