World leaders did not resolve bigger fights over AI at this week’s G7 summit in France. But in an official statement, they did agree on one smaller point: children using AI chatbots is its own safety issue, not just another version of social media. The news comes as a recent survey suggests nearly three-quarters of U.S. teens have used AI companions (chatbots used like a friend, coach, or romantic partner, et al.), and as the FTC asks major AI companies how their tools affect young people.

Civil courts are starting to ask more detailed questions about AI. These include: 1) if and when privilege applies when a party uses AI to prepare their case; and 2) whether uploading case materials to popular AI tools violates discovery-handling rules. For prosecutors, these early disputes are useful because they show where AI is already creating pressure on ordinary case-material controls. The chart below collects six early civil rulings already on the board.

San Francisco shut off a regional fusion center’s access to the city’s Flock camera data following improper outside-agency searches. (San Francisco Chronicle)

3D printers sold in New York could eventually have to block gun-part print jobs under a first-of-its-kind law. (AP)

Stockton, California, a city of about 325,000 east of San Francisco, has approved a multimillion-dollar plan to send drones to some 911 calls before officers arrive. (Stocktonia)

China is making a broad policy push to make AI a bigger part of what Chinese consumers use and buy, including humanoid robots. (Reuters)

The government of Estonia will look more closely at how to give AI agents digital identities. (Bloomberg)

An AI image-generation company is moving into medical scanners, while the company behind Snapchat is making AI-powered smart glasses. (The Verge / BBC)