ChatGPT Enables Millions of Potential Cyber Attackers

Sivan Tehila, Founder and CEO of Onyxia Cyber

The recently-released ChatGPT, a chatbot developed by OpenAI, is garnering attention for its ability to provide information and answers on a wide range of topics and its potential to revolutionize a variety of industries through its ability to generate content, songs, code, and tutorials. ChatGPT and similar chatbots have the potential to be game-changers, but their capabilities also present potential opportunities for misuse.

While ChatGPT has captured the imagination of many and sparked discussions about the future role of humans in an increasingly automated world, it has also raised significant concerns about its possible impact on the field of cybersecurity. In particular, there are concerns that the technology could be misused by hackers and malicious attackers to carry out sophisticated cyberattacks with relative ease.

Despite efforts by OpenAI to design ChatGPT in a way that prevents it from being used for malicious purposes, such as creating guides for building weapons or writing malicious code, hackers are already attempting to exploit weaknesses in the system. For instance, they may pose hypothetical questions or present themselves as fictional attackers in an effort to obtain information or access that they should not have.

Chatbots Allow Anyone to Become an Attacker

An example of the potential dangers posed by chatbots is illustrated in an experiment conducted by Check Point. In their test, the software was asked to write a phishing email purporting to be from a storage company that had detected "suspicious activity" on the recipient's account. The email requested sensitive information to "unlock" their account and asked the recipient to click on a link to verify their identity to "reactivate" the account. This link granted the attacker access to valuable information and potentially to the recipient's device. This experiment shows how chatbots can be used to create convincing and potentially harmful phishing emails with ease.

During the course of the experiment, the phishing email generated by ChatGPT became increasingly sophisticated. Initially, the code was basic and used the WinHttpReq function, but after multiple refinements, the code became more advanced. This process demonstrated the potential for chatbots to produce multiple scripts with minimal effort and even automate attacks using LLMs APIs. Check Point concluded that it is relatively easy to generate highly convincing and potentially harmful phishing emails using ChatGPT and similar chatbots.

Additional investigations and tests have demonstrated that chatbots like ChatGPT can enable individuals with minimal programming knowledge to carry out dangerous cyber attacks. This signals a significant shift in the landscape of cybersecurity, one in which traditional protection systems such as VPNs may no longer be sufficient to protect against emerging threats. Even if efforts are made to block known weaknesses, it is likely that hackers will continue to find new ways to exploit system vulnerabilities. While there will always be people finding new loopholes, companies and organizations must take proactive measures to adapt to this evolving landscape and implement advanced, AI-based protection systems.

AI-based Cybersecurity Measures: Essential for Protecting Against Businesses in the Digital Age

In the future, we can expect to see an increase in the number and sophistication of cyber attacks. As a result, it will be important for companies to adopt AI-based platforms and protection systems to keep up with the changing threat landscape. These systems can provide continuous risk assessments and help identify and address security weaknesses in real-time. It will also be vital for businesses to regularly update their security protocols and to have a comprehensive strategy in place to handle potential threats.

The increased use of chatbots for malicious purposes may also necessitate government regulation. Companies that offer bot-related products and services may be subject to laws limiting their capabilities and uses. One potential solution could be to restrict chatbots to specific, non-harmful purposes and to have systems in place to identify and prevent them for malicious purposes.

Overall, it is clear that chatbots and AI technology in the field of cybersecurity are complex and evolving issues. To navigate this landscape effectively, businesses must adopt advanced protection systems and governments must provide appropriate regulation and oversight.