The new AI scene can be very exciting, and is!
But the questions has to be asked, how much easier does this make attacks for threat actors?
AI has been used by threat actors to hack into corporate and disrupt business operations for many years now, but with the introduction of brand new generative AI tools such as ChatGPT and others, this also leaves us with brand new challenges for businesses everywhere.
Let’s consider three different scenarios in which AI/ChatGPT is used to help a threat actor carry out their attack:
1. A hacker can use a tool such as ChatGPT to generate personalized spear-phishing messages based on the target company’s marketing materials, CEO, other employees, relevant information, sponsorship, etc. ChatGPT will make it harder to detect phishing attempts as a whole and it’s important that we stay vigilant and have secure policies in place so we remain prepared.
2. An AI bot could make a phone call using a deepvoice to impersonate someone highly ranked within a company, such as the CEO or other highly ranked position to encourage an employee to do something that could potentially leave the company vulnerable, or give away important company information. Employees would never normally do these actions but since they are under the belief that this is what the company wishes, they do it anyways.
3. The impact of AI and it’s ability for threat actors to quickly create and develop more sophisticated attacks may always leave us catching up and shouldn’t be understated. AI can be trained using a large amount of phishing email data to create more realistic and convincing phishing emails, as well as other services such as live chats, phone calls, etc to sound as if it is a real person on the other side of that phone or screen. Attacks in general will be easier to automate, make sound more convincing, and continue to evolve leaving us in security to catch up.
AI is something that will take the world years to adjust to, and is seemingly going to be really hard to get ahead of, this is why it’s as important as ever to be aware of the tools that we have available and to include AI as a part of our threat modeling. I don’t want to take away from the general ‘hype’ around the new AI tools and their capabilities as I’m as excited as anyone, but it does leave us with a lot of new questions as far as security is concerned.
– Michael Walz