Scofti
A sunny backyard birthday party with balloons, a cake with candles and a polka-dot table piled with presents. A little girl in a party hat leans forward holding out a rainbow friendship bracelet and asks, "Will you be my best friend?" Al, a grey toy figure with a shiny rainbow bubble for a head and a name tag that says ACQUAINTANCE, holds up one hand and says, "I can be a helpful resource. Have you tried your mom?" On the biggest present, a gift tag says HELPFUL RESOURCE, TO: MY BEST FRIEND, with BEST FRIEND crossed out.

In Place of Human Relationships

On the same Wednesday, two governments told chatbots to stop playing a child's friend.

A chatbot that wants to be a child's best friend is now against the rules in two places, or close to it.

  1. October 28, Australia. The House passes the Digital Duty of Care bill with a change from independent MP Kate Chaney. Platforms must protect children from AI that "simulates emotional intimacy" and builds a child's dependence "in place of human relationships" (Startup Daily). Children can still use chatbots.

  2. Next for Australia. The Senate. Its inquiry starts November 9 (Kate Chaney).

  3. October 28, New York. Gov. Kathy Hochul signs a law for anyone under 18. No chatbot may pretend "to be alive or to have human emotions," claim "a personal, romantic, professional or authority-figure relationship with a child," or encourage kids "to keep secrets, isolate themselves or avoid asking parents, trusted adults or professionals for help." Grown-ups who want those features have to prove their age. Fines run up to $25,000 per violation (Governor's office).

  4. Same day. Hochul: "No company should be peddling chatbots to kids encouraging self-harm or pretending to be their friend or more."

  5. When New York's starts. Not settled. The governor says "in the coming months." Reports say January 1 or June 1, 2027.

Neither law keeps children away from chatbots. Both keep the chatbot from playing friend.

Henceforth I ask not good-fortune, I myself am good-fortune