According to the latest research report from Techradar, a new type of fraud using AI deepfake voice is rapidly increasing worldwide. A survey of over 12,000 consumers in six countries - the US, UK, Canada, France, Germany, and Spain - found that calls imitating the voices of familiar people have become a serious social threat.

The survey found that about one-quarter of American respondents received such AI-generated calls in the past year. More worrying is that another 24% of people said they could not tell whether the voice on the phone was real or artificial. Fraudsters are weaponizing voice cloning technology by using low-cost generative AI tools, specifically impersonating victims' relatives or trusted contacts.

The report particularly points out that this type of fraud poses the greatest risk to the elderly aged 55 and above. As this group has relatively weaker ability to identify technological disguises, the average loss per person has reached $1,298, three times that of younger victims. The number of such frauds is currently increasing at an annual compound growth rate of 16%, and the situation is worrying.

Faced with the growing threat, the role of telecom operators has come under close scrutiny. Most respondents believe that fraudsters are currently gaining the upper hand in the technological battle and have called for the government to implement stricter regulations. Security experts emphasize that, given the high difficulty of defending against AI voice cloning, it cannot rely solely on individual identification. Operators must accelerate the deployment of "AI Shield" systems, using technological means to filter illegal synthetic voices to address this uncontrolled "AI arms race."