Criminals are hacking into children's social media accounts and cloning their voices using AI to trick their parents into sending them money, MailOnline can reveal.
Even the most basic scammers are using simple AI tools online to turn just three seconds of a child's voice into deepfake audio files with an 85 to 90 percent match, security experts warn.
If more effort is put into cloning, hackers could even achieve a 95 percent vote match, according to research from security firm McAfee, putting parents at risk of being exploited by ruthless fraudsters.
The more accurate the clone, the greater the chance that a criminal will take advantage of a friend or family member's vulnerability and trick them into handing over their money.
It comes as fraud experts have issued new warnings about the latest 'hi mum' texts in which scammers 'prey on our goodwill with emotional stories' and dupe parents into thinking their children are in trouble.
Vonny Gamot, head of Europe, Middle East and Africa at McAfee, told MailOnline that AI is “becoming a catalyst” for online fraud, such as the “hello mom, hello dad” scam.
She said: 'The AI wave is going to give a new lease of life to everything that has been around for a while. The only thing is that we will see more and those AI threats will be even more deadly.”
Ms Gamot revealed how an increase in the number of young people using social media – and preferring to send voice messages over text messages – could be behind the rise in this type of fraud.
Hackers are targeting children's social media accounts, such as on Facebook, Instagram, Snapchat and TikTok, to obtain samples of their voices from videos posted online or from voice messages.
Ms. Gamot explains: “Hackers are going to take their voice out of social media. They're not going to take more than three seconds and create a new scam.”
She added: “You know this voice – and the reason you know this voice is because this voice has been sampled. The monster is real, but the AI used it to create a complete scam.'
The hackers then send voicemails or messages with the preview to the “mom” and “dad” contacts stored on the children's phones.
Ms Gamot explained that when a mother or father next hears their child's voice, they are likely to panic at that moment. “You lose your guard, you don't use your common sense,” she added.
But it's not just children who use their voices more. A McAfee survey found that 50 percent of adults share their voice data online at least once a week on social media or through voice memos.
Criminals hack into children's social media accounts and clone their voices using AI to trick parents into sending them money (file image)
The McAfee investigation shows that scammers are using AI technology to clone voices and then send a fake voicemail or call the victim's contacts, pretending to be in distress.
About 65 percent of adults were not confident they could tell the cloned version from the real thing.
The costs associated with AI voting fraud can be significant: 40 percent of people who have lost money say it has cost them more than £1,000, while 6 percent have been scammed out of £5,000 to £15,000.
Ms Gamot said her company is already identifying more than 1.5 million threats per day that are AI-based – with the 'hello mom, hello dad' scam text being a major form of fraud.
'Scammers will use AI tools that they can easily find online or on the dark web. They will taste something,” she added.
“In that case, it could be a vote that they're going to use against you, and they're going to try to get data out of you so they can resell it, or they're going to try to get money out of you. In both cases you are not in good shape.'
Ms Gamot said the “digitalisation of the younger generation is becoming increasingly important” – with children gaining access to devices at a younger age and schools asking children to do homework online.
McAfee advises victims of potential AI fraud to first “take a breath” and then call their friend or family member to make sure they are not in trouble.
Ms Gamot explained: 'We're really starting to see these threats increase and we want to make sure we give everyone the tool to actually protect themselves as they continue to enjoy their digital lives online.
“The only thing is we want to make sure people can do this in a safe way.
'And before we try to sell our solutions, we also want to share a few tips that we think you know, that don't cost anything, and are actually quite efficient, and really start by taking another deep breath and thinking twice think what is it about?'
Above are examples of scammers posing as children to trick parents into sending money. There's a resurgence of the 'hello mom, hello dad' scam – increasing with the development of AI
McAfee also suggests setting a verbal “code word” for children or trusted close friends that only they can know if they are in trouble.
Scam expert Nick France from leading cybersecurity company Sectigo also fears that this type of fraud will only increase in the next five years, especially due to the increasing use of AI deepfakes.
Mr France told MailOnline: 'The ever-increasing sophistication of AI has made it possible for cybercriminals to successfully imitate someone else's voice and successfully impersonate them for a number of attacks.
“People think phone scams, which successfully manipulate someone's voice, are a mission impossible, but the reality is that AI deepfake voice technology is more democratized than we like to think. It doesn't take an MIT graduate to make this happen.
Just last month, Sadiq Khan was the victim of a deepfake audio in which he said: 'I don't care what Remembrance weekend is.'
A McAfee survey found that 50 percent of adults share their voice data online at least once a week on social media or through voice memos. (file image)
The rise of AI voice clones has led to a resurgence of the “hello mom, hello dad” scam, which first appeared on WhatsApp in 2021.
It has since grown on other platforms such as text messaging and now voice messaging.
Data shared by TSB with MailOnline shows that 'friends and family fraud' is responsible for 53 percent of all impersonation, with 93 percent of that coming from WhatsApp.
The average loss for the victim of fraud is over €1,600.
TSB fraud spokesman Matt Hepburn said: 'Friends and family impersonator scammers prey on our goodwill with emotional stories and pleas for urgent financial help, simply to steal money meant for someone close to us.
'If you receive one of these texts, please contact the person directly before proceeding, and certainly before ever making a payment, as it is most likely a scam.'
Chris Ainsley, head of fraud at Santander, told MailOnline: 'It's normal to want to take immediate action on messages from loved ones asking for help, but scammers will pose as friends and relatives to scare you into transferring money without thinking about what you're doing.
'Always take the time to think about what is being asked of you. Any request to transfer money to a new account or excuses as to why someone cannot speak directly on the phone should be treated as a red flag.”