Cybercriminals have been seen using AI chatbots like ChatGPT to develop phishing websites and enhance their chances for online fraud and scams. A team of researchers at the University of Texas at Arlington has developed software to help prevent these AI phishing scams. In this post, we will dwell in detail on how these researchers have created this software to prevent AI phishing scams.
Source - Tech Explore
This software helps an AI chatbot detect and reject any suspicious instruction prompts that can be used for scams or developing phishing websites. Although AI chatbots already have some in-built detection features that detect the prompt upon engagement, these have certain loopholes that can be exploited by attackers.
If the prompt is inserted with a bit of planning, the attackers can easily bypass this detection and use the AI chatbots to attack and create phishing websites. The team of researchers included Dr. Shirin Nilizadeh and her doctoral students. They observed the AI chatbots closely and found the loopholes that were used in the attacks.
Dr. Nilizadeh stated, "These tools are very powerful, and we are showing how they can be misused by attackers."
After the group noticed the loopholes, they then identified various prompts that were used by attackers to develop phishing websites. They leverage this knowledge and train their software accordingly. Enabling it to identify and react to specific keywords and patterns of the attacker’s prompts. This enhanced the ability of the software to detect and block these prompts and any suspicious prompts from the AI chatbots.
With the introduction of ChatGPT by OpenAI, interest in LLMs surged and additional chatbots were introduced by many companies, including Google’s Gemini, Microsoft’s Copilot, and many more. Dr. Nilizadeh and her team also reached out to these tech behemoths with the proposal of integrating their research and software into new AI security measures to improve the cybersecurity aspects related to artificial intelligence.
Suggested Read: Lessons for Developers from the Largest DDoS Attack in History
This software gained immense popularity when first introduced, and the team's work was heavily appreciated and highlighted in a recent publication at the IEEE Symposium on Security and Privacy that received the Distinguished Paper Award. The cybersecurity community has also shown support and great interest in the work of this team.
Conclusion
We have seen a gradual increase in the number of cyberattacks in the digital world, with each attack being more evolved and unable to be traced back. This shows how much potential these attackers put into these attacks and how they keep getting better and better with technological advancement.
One such advancement included artificial intelligence, and with new AI chatbots like ChatGPT, many malicious developers were able to create phishing websites. The cybersecurity community will benefit from the team of researchers at the University of Texas who built this software to prevent AI phishing scams. It will be a beneficial addition to the AI security framework in the future. What do you think about this software? Let us know your thoughts in the comment section below.
ExcelliMatrix provides multiple services, in addition to software development support, we are also famous for our cybersecurity support. Our team of experts will guide you or your business through the rigorous cybersecurity process. With our latest cybersecurity measures, you and your business will both be protected. For a free consultation, please contact us at 406-646-2102 or email us at sales@excellimatrix.com.
Stay connected with us on LinkedIn and Facebook, and follow us on Twitter for more information like this. You can also subscribe to our weekly newsletter for more technology and security information.
