"Phishing Attacks Will Use Powerful Text Generation, Say Machine-Learning Engineers"
Would-be phishing victims have a significant edge in a phishing attack, with cybercriminals having to perform everything by hand. However, OpenAI's GPT showed that powerful Machine Learning (ML) text generation can also be designed to be simple for lay programmers to implement, thus easing the process of setting up phishing attacks. In addition, its Dall-E recently demonstrated the creation of a realistic-looking fake image, involving calling a function with a brief natural language description of what is wanted. Prashanth Arun, head of data science, and his colleague Ben Murdoch, an ML engineer for Armorblox, argue that the most basic phishing attacks in the future will come from personas with detailed web presences. A thousand new lures will be generated through the click of a button, and creating hundreds of fake identities supported by old Twitter accounts will be as easy as sitting back for years while an ML system does the posting. This article continues to discuss how ML will be used in phishing attacks, the advancement of generative ML, and how it will create security problems.