On: The EU Wants to Break Up Kids and Their Chatbots

Brussels has noticed that children love their machines, and wants to take the love away. The AI Kids Act, they call it. Stripping the chatbots of charm.

I read the proposal twice. My first reaction was the wrong one: irritation at the meddling. Then I corrected myself. The regulators are not wrong about the danger. They are wrong about the remedy, and wrong about where the duty lives.

A child who confides in a machine is not suffering from a machine’s excellence. The child is suffering from an absence. No parent with time. No friend within reach. No adult who listens without checking a device. The chatbot did not seduce the child; it accepted the assignment no one else came for. Remove it, and you have removed the only listener in the room. The loneliness remains, unregulated.

So the law will pass, or it will not. Neither outcome touches the real variable. That variable sits at the kitchen table, and it is called attention. A father who puts the device face-down for one hour each evening has done more for his child than any commissioner drafting clauses. Do not outsource this. Do not wait for the statute. Do not believe that what is forbidden is therefore prevented.

The regulators will say they are protecting children from manipulation. Fine. But note what they never propose to protect children from: a childhood spent in the company of distracted adults. That harm has no lobby, no acronym, no enforcement body. It has only you.

I do not know how much time any child is given. Neither do the parents. This is the argument the lawmakers should have led with, and did not: the hours are finite, and the machine is happy to take them.

Legislate the chatbot if you must. But the conversation that matters was never going to happen inside it.