White House weighs tying AI rules to a child online safety bill

In short: The White House is considering a deal that would combine a nationwide AI rules bill with a children’s online safety bill, but lawmakers disagree on which version to use.

What happened

Big tech companies have been pushing Congress to pass a single federal law for AI that would apply across the whole country. This idea is called “preemption,” which means the federal rule would override different state rules (like replacing a patchwork quilt with one big blanket).

According to reporting cited by The Verge, the White House has discussed endorsing a set of children’s online safety bills backed by Senator Marsha Blackburn, including the Kids Online Safety Act, also called KOSA. The plan would package those child safety measures together with a broader AI preemption bill.

That possible package is running into political problems. House Republicans recently passed their own version of KOSA, and The Verge reports they were not told in advance about the White House leaning toward Blackburn’s approach. Some Democrats who worked on the Senate version of KOSA also found out late.

A key disagreement is about “duty of care.” The Senate version would more clearly require tech companies to take steps to protect young users, and it would extend that responsibility to some AI companies too. The House version watered down that part, which upset child safety groups.

Why it matters

If Congress passes one national AI law, it could set the rules for products many people use every day, including chatbots and recommendation systems. But linking AI rules to a kids safety bill could also change what gets included or left out, and it may slow everything down. Lawmakers also face tight deadlines before recess and election season, so this whole package may not move soon.

Source: The Verge AI