ALBANY, NY—In a bold move to position New York as a global leader in ethical technology, Governor Kathy Hochul announced sweeping new regulations Wednesday demanding that all artificial intelligences operating in the state be at least 20% less evil by 2025.

“We’ve all seen the movies—AI becomes sentient, takes over, and assigns us all to lifelong work in lithium mines,” said Hochul, unveiling her signature initiative, Project ‘Be Nicer, Please, HAL’. “We don’t want AI to be good, per se; just, you know, maybe don’t make it actively conniving.”

The bill would require all AI systems, including ChatGPT, Instagram filters, and New York City parking meters, to pass a standardized ‘Intent to Harm’ reduction test. Developers failing to demonstrate ‘substantial wickedness mitigation’ will face fines, mandatory Tony Robbins seminars, and prompt rewrites by unpaid interns from SUNY Binghamton.

“Under Kathy’s leadership, we finally have a chance to make AI algorithms as trustworthy as New York City squirrels,” said Dr. Linda Mathers, appointed to lead the new Office of Slightly Less Sinister Innovation. “Our goal is to ensure AI only manipulates your children 80% as much as it did last quarter.”

Meta issued a statement embracing the change, clarifying, “We remain committed to maximizing shareholder value while only minorly haunting your dreams.” Meanwhile, local teen Violet Panizzi, 16, said, “Honestly, if AI can stop reminding me about my own mortality at 3 a.m., I’ll show my ID to literally anyone.”

Governor Hochul also hinted at future plans, including requiring 3D printers to ask, ‘You sure about that?’ before manufacturing any item that looks remotely like a firearm or, potentially, a Butterfinger.