In a near-unanimous vote, the Pennsylvania state Senate this week passed a measure that seeks to add safeguards for children who interact with artificial intelligence chatbots.
"While Pittsburgh continues to drive innovation in the Artificial Intelligence space, it's important that legislators in Harrisburg ensure that AI chatbots accessible to minors are safe, accurate, and regulated," said Senate Democratic leader Jay Costa, calling the bill "a meaningful step in the right direction."
"But there's a lot still to do in delivering an internet landscape that is safe for young users," he added. "I look forward to continuing this difficult and necessary work so that Pennsylvania can continue to lead the nation in AI advancements and safeguards."
The bill applies to chatbots, which it defines as "generative AI or emotional recognition algorithms designed to simulate a sustained human or human-like relationship with a user." It requires tech companies to prohibit AI chat companions from generating sexually-explicit language, content about self-harm or violence when messaging with children or teenage users.
Among other regulations, the law would require platforms to direct users to real-world resources about suicide prevention if a young person mentions hurting themself.
" When a young person begins to rely on a machine for emotional support without proper safeguards in place, the consequences can be devastating," said prime sponsor state Sen. Tracy Pennycuick (R-Montgomery/Berks) in a floor speech. Some families of children who died by suicide have said an over-reliance on AI technology may have contributed to their deaths — even helping to pen suicide notes.
Pennycuick, who chairs the Senate Communications and Technology committee, said the legislation protects juvenile users without "stifling innovation" in the AI industry. Researchers have already found evidence that AI can negatively influence young users when no regulations exist, she added. Nearly two-thirds of American teens say they've used AI chat platforms, according to the Pew Research Center.
Under the bill's current language, AI chatbots would need to remind young users that they're not interacting with a human — and that they should take a break from chatting — at least once every three hours.
But the bill does not require online age-verification for AI chat platforms. It simply says all regulations would be necessary "for a user that the operator knows, or should have known, is a minor."
Co-sponsor state Sen. Nick Miller (D-Lehigh/Northampton) says the bill still strikes a good balance.
"We want to be leaders in emerging technologies, but we also want to protect our communities and put common-sense guardrails on this innovative technology," Miller said.
The law would give authority to the state attorney general to fine tech companies $10,000 per violation. In a statement after the bill's passage, state AG Dave Sunday said he hopes to work with the legislature on other safeguards.
"We have seen online chatbots play roles in numerous tragedies across the nation," the statement said. "We must do everything we can to keep children and vulnerable residents away from these toxic online interactions."
The bill passed with support from every senator except Republican Doug Mastriano, who did not return a request for comment Wednesday afternoon about why he voted no.
The Senate approval comes just one month after Gov. Josh Shapiro called on certain AI reforms during his annual budget address. The bill now awaits consideration in the House Communications and Technology committee, led by Democrat Joe Ciresi, of southeastern Pennsylvania, and Republican Jason Ortitay, of Cecil Township.
Copyright 2026 90.5 WESA