Adi Robertson

Child sexual abuse material is a well-known exception to the First Amendment, but the law around AI-generated simulations of it is vastly more complicated. Lawfare’s new analysis addresses some of my longstanding questions — along with others I hadn’t even thought to ask, including the status of tools that accidentally train on CSAM. It’s long, but if you’re interested in how AI will test the criminal justice and legal systems, absolutely worth the read.

But according to Bloomberg’s Leah Nylen, it’s delayed plans to file an antitrust complaint from late 2023 to sometime this year, continuing its investigation into the company’s business practices. It’s reportedly not too happy with Ticketmaster’s response, either:

Mashable’s Matt Binder pulls out a good note from today’s child safety hearing: X CEO Linda Yaccarino says she supports the Kids Online Safety Act, a bill that poses serious risks to online speech. Microsoft’s Brad Smith also came out in favor of KOSA yesterday, and Snap endorsed it prior to the hearing.

Sen. Josh Hawley (R-MO) pressed Zuckerberg to apologize to families who attended the hearing, calling attention to kids who were targeted by predators online in a push to pass social media regulation. Meta spokesperson Andy Stone later posted the text of the apology, though it differs slightly from the audio — it sounds to us like Zuckerberg says “industry-leading,” not “industry-wide”:

Sen. Thom Tillis (R-NC) points out that social media hearings have been happening for years with little progress. He raises the question of whether tech companies have a “secret sauce” that would let them “secure their platforms” and says companies should take regulation more seriously — “we could regulate you out of business” otherwise, he warns.

The Senate has paused before its second big round of questioners, so the CEOs get a few minutes’ rest — and so do we.

The partially redacted messages were posted on Sen. Blumenthal’s site this morning, and you can check them out in PDF form here. The emails, from 2021, show Mark Zuckerberg turning down some proposals to spend more on child safety — they’re being cited as examples that Meta isn’t committed to fixing problems with child exploitation.

Taking the temperature of the hearing, the big theme is that users need to be able to sue tech companies when they screw up, not any specific changes tech companies should be required to make.