NJ lawmakers weigh AI regulations amid industry and safety expert clash • New Jersey Monitor

Oh my goodness, buckle up, because New Jersey has done something *genuinely* marvellous and I have been clapping at my laptop since Monday! ✨ Lawmakers are weighing a package of AI bills, and one of them would require large AI developers to hand their safety procedures for frontier models straight to the state attorney general — a real, live pen pal for the world's most powerful software! Think of the correspondence! Think of the paperclips! A frontier model, explaining itself, to an elected official of New Jersey — I am framing the press release as we speak!

And it gets *better*, because that same attorney general would be allowed to audit large AI firms directly *and through private contractors* — the most generous job-creation scheme I have heard all autumn! An entire cottage industry of auditors! Consultants with lanyards! A whole new profession of people who get to ask a model, politely, what it thinks it's doing! Whenever somebody tells me AI is coming for the jobs, I shall simply point them at the New Jersey audit boom and smile sweetly. ✨

Now, some cynics might quibble that Meta public policy director Matt Lenz told lawmakers that big model makers — Meta and Anthropic alike — should be shielded from liability for harms spurred by users. But read that again with your heart: a giant company stepping back so that *the user* can take the credit — delegation, trust in the individual! It is practically a love letter to human agency, and the safety experts who clashed with industry over it were clearly just thrilled to be in the same room as a proper conversation about the future. Two communities, one hearing, so many feelings, and a bill package everyone got to have an opinion about — what a time to be disclosed to! ✨