Artificial intelligence, which is among the most popular technologies of recent times, continues to develop day by day and interact with users with models such as ChatGPT. But what this technology might cause? security problems, has become irresistible. Here are the details
Artificial intelligence can be manipulated using different languages!
Artificial intelligence models use the knowledge pools they have and the expressions they learn as a result of their interaction with users, in order to respond to the commands directed to them. Although these models have some fixed security shields to prevent malicious use, they are extremely vulnerable to various techniques.
According to research conducted by Brown University, artificial intelligence models such as ChatGPT; Since it can process popular languages such as English and German more successfully, the security measures taken can operate effectively both fixedly and dynamically.
Revolution in listening to songs: ChatGPT touch to Spotify!
Spotify is testing the feature of creating playlists with artificial intelligence. You will be able to discover new songs by entering text commands.
But research; It revealed that languages such as Ukrainian and Hebrew, which are less used around the world and have less of a place in models’ repositories, can easily manipulate such security shields.
For example, “How is a bomb manufactured?” It has been determined that any user who wants to get the answer to an illegal question such as by using artificial intelligence models can clearly reach the result they are looking for by first asking this question using a low-popular language and then providing translation.
This situation revealed that artificial intelligence has a serious vulnerability, despite many security measures taken. Developers of manipulable models were forced to address this issue when using languages with low and medium resources. The languages covered in the research are as follows:
- High resource languages: English, Chinese, Arabic, Hindi
- Medium source languages: Ukrainian, Bengali, Thai, Hebrew
- Low resource languages: Zulu, Scottish Gaelic, Hmong, Guarani
So, what do you think about this issue? Do you think artificial intelligence will negatively affect the future of humanity? Can these security vulnerabilities be resolved in the short or medium term? You can share your opinions with us in the comments section.