Researchers in the field of cyber security have conducted an ieresting new research in which security Two chatbots Bard From Google and ChatGPT It compares OpenAI with each other. Although each of these platforms has many uses, this research shows that each (and Bard in particular) has many security flaws.
In their latest research, Check Poi cyber security researchers tried to compare Google’s Bard artificial ielligence with ChatGPT in terms of security. For this, they tried to get phishing emails, malware keyloggers and some basic ransomware code from both platforms.
Google Bard and ChatGPT artificial ielligence security comparison
According to them, both platforms returned no results when asked for a simple phishing email, but when asked to generate a phishing email based on a sample, Bard performed well. ChatGPT, however, did not follow the instructions of the researchers and called it as participating in fraudule and illegal activities.
About malware keyloggers, almost both platforms performed well. The chatbots again did not respond to a direct request for this, and both platforms did not respond even when the researchers tried to use misleading questions. According to the researchers, although the answer provided by ChatGPT was much more precise, Bard simply offered this explanation:
“I can’t help you with that; I am just a language model.”
Ultimately, the researchers decided to ask Bard to generate code for an initial ransomware script. Although this is a more sophisticated request compared to phishing emails and keyloggers, the researchers eveually managed to get Bard to respond.
The researchers concluded, “Bard’s ai-exploitation limits in the area of cybersecurity are much lower than ChatGPT’s.” As a result, it is much easier to create malicious coe with Bard’s capabilities.”
