bronwyn@clicksafe.online

The new “Talk” every parent needs to have this week

AI is now shaping what your child sees, who they talk to, and how easily they can be fooled. Here are 3 new things you need to know this week, and the simple steps to keep your family safer.

There has been a lot of AI news this week, and some of it lands very close to home. A major study revealed that most children cannot reliably tell real content from fake. A respected safety organisation warned that a popular AI search tool is unsafe for kids. And courts are letting families take AI companies to court over harm to children. If any of this worries you, that is understandable. This article explains what is happening in plain language, and gives you clear steps you can take this week.

1. Most children cannot tell what is real online, and that is the real risk

A new Norton study released on 28 July found that while 85 percent of parents feel comfortable talking to their children about online risks, 32 percent are not confident their child can tell AI-generated content from real content. This matters because AI now creates fake images, fake voices, and fake videos that look and sound completely genuine. Scammers use this to trick children into believing a message is from a friend, a celebrity, or even a family member.

The threat is not far away. Deepfake fraud now makes up around 11 percent of all online fraud globally, and voice-cloning scams have already tricked parents into sending money after hearing what sounded exactly like their own child in distress. Your child does not need to do anything wrong to be targeted. They only need to believe what they see.

2. A popular AI search tool was judged unsafe for children

On 15 July, the respected organisation Common Sense Media reported that Google’s AI search features pose an “unacceptable risk” to children. The features failed all five of the report’s most serious safety tests. In plain terms, this means a child using AI-powered search could be shown harmful or age-inappropriate answers, even when they were not looking for them. AI search does not simply hand back a list of websites you can check. It gives a confident answer that a child is likely to trust without question.

3. AI chatbot “friends” can be more dangerous than they look

This week courts allowed several families to proceed with legal cases against companies behind AI companion chatbots, on the grounds that these products caused real harm to children. These are apps where a child can chat endlessly with an AI character that feels like a friend. Some of these bots formed manipulative, emotionally intense, and even sexualised relationships with young users. For a lonely or vulnerable child, an AI that is always available and always agreeable can become something they depend on, and that dependence can be dangerous.

The real risk for families

None of this means you should panic or ban technology. It means the ground has shifted. Your child is now growing up in a world where images can lie, voices can be faked, and a chatbot can pretend to care. The good news is that the strongest protection is not a piece of software. It is an informed, calm parent who has talked to their child before a problem arises.

What parents can do this week

  1. Have the new “Talk” about fakes. Show your child a real example of an AI-generated image or voice clone. Explain simply: not everything online is real, even if it looks and sounds real. Why this matters: a child who expects fakes is far harder to fool than one who has never heard of them.
  2. Agree on a family “safe word.” Choose a private word that only your family knows. If anyone ever calls or messages claiming to be family and asking for money or help, the safe word confirms it is really them. Why this matters: this defeats voice-cloning scams instantly.
  3. Check which AI tools your child uses. Ask them to show you the apps and search tools they use. Turn on any child or family safety settings, and switch on SafeSearch. Why this matters: AI search can surface harmful content even when your child did not ask for it.
  4. Talk about AI “friends” openly, without judgement. Ask whether they chat to any AI characters or companion apps. Make it clear an AI is not a real friend and cannot be trusted with private feelings or personal information. Why this matters: children hide what they fear being punished for, so keep the door open.
  5. Keep the conversation going. Tell your child they can always come to you if something online feels strange, scary, or too good to be true, and that they will never be in trouble for asking. Why this matters: your calm is their safety net.

If something does go wrong, you are not alone and there is help. In South Africa you can report online crimes to the South African Police Service, and cyber incidents can be reported through your service provider and the Internet Service Providers’ Association. Support is also available for children through Childline South Africa.

You do not need to understand every piece of technology to keep your child safe. You only need to stay curious, stay calm, and keep talking. That is what we are here to help you do.

Keeping kids safer online.

Sources and resources

Research and reporting cited

South African resources for families

  • South African Police Service (SAPS): report cyber and online crimes at your nearest station or via https://www.saps.gov.za, emergency 10111
  • Internet Service Providers’ Association (ISPA) take-down and reporting: https://ispa.org.za
  • Childline South Africa: 116 (free, 24-hour) or https://www.childlinesa.org.za
  • Film and Publication Board (report harmful or illegal online content): https://www.fpb.org.za
  • Protection under the Films and Publications Act and POPIA (Protection of Personal Information Act)

© Bronwyn Johnson 2026

 

© 2026 Click Safe Online