OpenAI has introduced a new web search function in Chatgpt, designed to improve user consultation experience. This tool allows you to perform searches directly from the chatbot interface, offering an advanced way to access information on the web. However, Kaspersky experts have indicated that, as with traditional search engines, there are safety risks that users must take into account.
The integration of the web search in Chatgn promises a more fluid and efficient user experience. Users can obtain quick and precise answers without changing platform. However, Vladislav Tushkanov, manager of the Research Development Group in Kaspersky, warns that this functionality can also expose users to suspicious or phishing websites.
“While the search for chatgpt mainly provides links to official websites, in some cases it could also choose suspicious or fraudulent resources. On the other hand, the global knowledge of Chatbot could act as protection for recognized companies, since the underlying language model can have precise information about its official websites. However, this does not absolutely guarantee that phishing links in the answers are not filtered, ”he says.
An example of this risk was observed when a search page search for a cryptocurrency exchange showed a phishing resource in the results. Although the chat response body included legitimate domains, the presence of fraudulent links in the results panel highlights the need for caution.
Cybersecurity analysis: chatgpt
Kaspersky’s analysis revealed that, in general, searches for brands known as Google, Facebook, Amazon, Microsoft and DHL presented safe and legitimate links. This suggests that Chatgpt can correctly identify the official domains of recognized entities. However, the possibility that fraudulent links are filtered in the results still exists, especially in the case of searches related to cryptocurrencies and other sensitive resources.
Cyberamezas and risks of the new chatgpt web search function
Tushkanov explains that the cybernetic risks associated with the new chatgpt search function are similar to those of traditional search engines. Phishing links may appear temporarily in the first search results, which could affect the information that Chatgpt summarizes. Although the global knowledge of chatbot can act as protection for recognized companies, it does not absolutely guarantee that phishing links are not filtered into the answers.
Security recommendations
To mitigate these risks, Kaspersky experts recommend users to address chatgpt search results with the same caution as those of any search engine. The suggested security measures are included:
- Verification of links: Before clicking on any link, users must verify their authenticity. Fraudulent web pages usually resemble the originals, but present subtle differences and errors.
- Use of reliable links: It is advisable to save reliable links or manually enter the official services URL to avoid fraud.
- Integral Security Solutions: Use security solutions such as Kaspersky Premium, which protects cryptocurrency wallets and warns of suspicious websites and phishing threats.
