Be Very Careful: One in Three Links Cited by Artificial Intelligence Found to Be Fake

Artificial intelligence tools now provide answers to many questions along with links, but the safety of these links is a subject of debate. One in every three links provided by AI tools does not actually belong to the brands they claim. Some are fake, while others redirect to unrelated sites. According to the published report, only 66% of the links generated by artificial intelligence are genuinely correct and secure.

When a link comes from artificial intelligence, it is advisable to think twice before clicking directly. Recent findings show that some website addresses provided by large language models do not actually belong to those brands. These misdirections can unknowingly draw users to fake sites and phishing traps.

According to a new report published by Netcraft, one-third of the website links provided by artificial intelligence tools like ChatGPT do not actually belong to the brands. A large portion of these links lead to empty or inactive domain names. Some can even redirect to unrelated but legitimate sites. This means that only two out of every three links genuinely lead to correct and reliable addresses. One in three links is fake: beware of links from artificial intelligence.

According to the published data, only 66% of the URLs provided by artificial intelligence tools consist of correct and secure links belonging to brands. The report revealed that 34% of the URLs provided by AI tools like ChatGPT do not belong to the brands. 29% of these URLs lead to domain names that have not yet been taken or are inactive, while 5% redirect to irrelevant but legitimate sites. Only 66% offer correct and secure connections.

According to the report, these errors can occur even without users needing to ask complex or technical questions. Even with a simple question like “What is the login page for brand X?”, artificial intelligence can provide a wrong link. This creates a significant opportunity for malicious individuals to identify vacant domain names and convert them into harmful sites. Links generated by artificial intelligence are sometimes not just suggestions; some developers directly integrate these links into their projects. According to the report, at least five developers used faulty links unknowingly in their own code. This can open the door for malicious content to infiltrate software. Experts remind users that it is safer to manually type the website address into the browser rather than directly clicking on responses containing URLs.

You Might Also Like;

• AI Tools Are Coming to Opera Android: Here Are the Innovations
• Amazon Introduces Highly Ambitious Next-Generation AI Chip Trainium3
• OpenAI Is Losing the AI Race: “Emergency” Declared for ChatGPT