At Kaspersky, we continue to investigate cutting-edge technologies and study their potential impact on both cyberattackers and defenders.Ĭuriously, in the comments to my previous post about Orenburg, a dear reader suggested we visit her nearby hometown of Gui (pronounced Guy not Gay). (iii) ChatGPT is not yet ready to radically change the cybersecurity game, at least as far as phishing detection is concerned.
(ii) But it has significant limitations for #cybersecurity due to “hallucination” tendency, limited context window, randomness of output, as well as possible prompt injection attacks. (i) For a zero-shot system the results are amazing – it is hard to argue with, even for a die-hard sceptic. So we conducted an experiment to see a practical example of how an LLM can be applied to a real, albeit simplified, cybersecurity task, and describe all the interesting things we learned along the way. What if all our complex, multi-layered detection systems are now obsolete, and a single LLM (large language model) to rule them all can fight cybercriminals better than we do? How good ChatGPT is at recognizing overtly malicious links?Īt Kaspersky we work hard on applying machine learning technologies to #cybersecurity tasks, specifically models that analyze websites to detect threats such as phishing. Here is a valuable advice how to do that (+ bonus track: the dark side of the moon) 👉 Given this - two questions arise: is it possible to get rid of LinkedIn fakes, and how can you protect your company’s brand? But one thing they have in common is that they don’t give a hoot for the HR-brand or the reputation of the companies where they supposedly work. For instance, in the fall of 2022, security expert Brian Krebs uncovered a whole bunch of fake LinkedIn accounts purporting to belong to the Chief Information Security Officers of various major international companies. The downside of this is the relative ease of creating plausible fake profiles. The platform is designed for communication among professionals, which automatically implies contact with new people, almost complete transparency of user information, as well as a fairly high degree of trust in total strangers. FAKE ACCOUNTS ON LINKEDIN: TIME FOR A PURGE!Īmong social networks, LinkedIn holds a rather unique position.
0 kommentar(er)
