OpenAI has testified in support of Illinois SB 3444, which would protect frontier AI developers from lawsuits over "critical harms" — defined as 100+ deaths, serious injuries to 100+ people, or $1B+ in property damage — as long as the lab did not intentionally or recklessly cause the incident and had published safety reports. The bill applies to any model trained on more than $100M in compute, covering all major U.S. AI labs. AI policy experts describe it as a more extreme liability shield than anything OpenAI has backed before; polling shows 90% of Illinois residents oppose the exemption.