DeSantis’ AI Bill of Rights clears Senate — but House won’t touch it

The Florida Senate overwhelmingly approved the DeSantis-backed “AI Bill of Rights” Wednesday — but it won’t matter because the Trump-aligned speaker won’t bring the bill to the House Floor.

“A bill that that hasn’t moved in the House is not going to be brought up at this time,” Speaker Daniel Perez, a Miami Republican, told reporters. He was responding to a question on a medical freedom bill, but the statement would apply to multiple bills that — one week before session is scheduled to end — haven’t budged in the lower chamber.

“The White House’s position on AI and the House’s position on AI … are on the same page,” he said. “We do believe that the federal government should take care of AI and in whatever legislation or policy has to pass on a national level as opposed to doing it on a state-by-state basis.”

The Senate passed the measure anyway. Only two lawmakers — Republican Sens. Erin Grall of Vero Beach and Don Gaetz of Crestview — voted against it.

It’s yet another bill that’s fallen casualty to a relationship so tense between the House and Senate that budget negotiations have hardly progressed. Discussions are moving so slowly that politicos are quietly considering either extending the regular session or calling a special session to complete the work. A budget is the only package legally required to pass each year.

SB 482, sponsored by Republican Sen. Tom Leek, was strongly pushed for by Gov. Ron DeSantis. For months, he urged the legislature to put limits on artificial intelligence’s power, warning of an “age of darkness and deceit” if it’s not constrained. This put him squarely opposite President Donald Trump, who’s allied himself with tech titans like Elon Musk and Mark Zuckerberg.

Trump even signed an executive order intended to largely prevent states from passing AI regulations. His administration reached out to Perez to talk about the Florida bill, the Daily Signal reported.

SB 482 would have banned companion chatbots — AI systems that mimic emotional connection — from speaking to minors without parental consent, and require bots to remind users they are not human.

“There’s an inherent evilness when we allow machines to create and sustain a relationship that a user believes to be real,” Leek said from the Senate floor Wednesday. “That evilness is only magnified when that machine interacts with a child or vulnerable adult.”

AI companies that violate the bill would have a 45-day “cure period” to fix any mistakes, according to a new amendment to the bill approved Wednesday. If they don’t, or if the attorney general deems their violations too egregious, they could face $50,000 fines, the Phoenix previously reported.

The platforms could have to pay up to $10,000 to a minor it recklessly allows onto its server without parental consent.

Companion chatbots wouldn’t include software used primarily by businesses, theme parks, or “artificial intelligence instructional tools,” used in schools, the bill says.

A parent would have the ability to opt-out their child from using AI tools at school. Elementary schools would be banned from providing access to AI unless school personnel supervised the use, it is for translation support for English learners, or for disability accommodations.

Parents could have had the ability to opt-out of their child using these AI tools.