ChatGPT offers automatization and potential business gains as well as presenting unique risks. Yet, with the same technology, hackers can use ChatGPT to manipulate victims, carrying out more sophisticated attacks and hence make it harder for users to detect potential cyberattacks.
According to NordVPN research, on dark web forums, the number of threads on ChatGPT rose from 37 to 91 across April, indicating that bot exploitation has become a popular topic among dark web surfers. This leads to the possibility that ChatGPT and social engineering attacks such as phishing, smishing or pretexting, could be on the cards. This is based on the potential for the technology to make things easier to steal valuable information.
These concerns relate to ChatGPT being able to generate convincing dialogues or scripts, and hence carry the potential to manipulate people into revealing confidential information or performing certain tasks.
A related concern is with cyberattacks becoming more sophisticated, since poorly worded emails and grammatical errors can be easily avoided with the use ofChatGPT. This allows people from countries where English is spoken rarely to participate in cybercrimes on a greater scale and achieve higher success rates.
Those with criminal tendencies using the dark web can obtain expertise on such topics as “How to break ChatGPT,” “ChatGPT jailbreak 2.0,” “ChatGPT – progression of malware,” and “ChatGPT as a phishing tool.”
Leaking sensitive information
A different concern is with the security of ChatGPT itself. In March 2023, ChatGPT experienced a data breach. This exposed users’ personal information, such as chat logs and credit card details. Such events prove that using this platform creates high risks to privacy.
This means that any kind of information added to the chatbot may be at risk of being leaked, especially when ChatGPT or other AI-powered tool is used for marketing purposes or writing emails.
Prevention is always cheaper than a breach
There are some important principles to practice in order to protect individuals and businesses. These include:
Educate employees
Social engineering attacks are becoming increasingly sophisticated. Thus, even more attention to employee training is needed.
One of the most effective ways to do that is to organize cyberattack simulation exercise and openly discuss the results.
Have a business resilience plan
Risk assessment is a must for any company, no matter the size or industry. Once a company knows (and acknowledges) its risks, it can be useful to create a plan to:
- Tackle any challenge that might arise.
- Invest in prevention, detection, and threat mitigation.
- Establishing quality cybersecurity tools that prevent threats and offer network segmentation, identification, and access management.
- Maintaining IDS policies and data integrity checks.
- Putting in place clear post-mortem analysis and backup policies.
ChatGPT is provides useful technology helping us automate tasks and increase efficiency. However, as most chatbots are available to everyone, the same goes for bad actors that can use this technology to conduct sophisticated attacks that the human eye can no longer detect.
