Illinois is becoming a proving ground for frontier AI liability policy after OpenAI and Anthropic backed opposing bills in the General Assembly. OpenAI supported SB 3444, which would limit developer liability for catastrophes involving death or serious injury to mass victims, while excluding some scenarios involving chemical, biological, radiological, or nuclear weapons. Anthropic opposed SB 3444 and backed SB 3261 instead, which would require developers to publish public safety and child protection plans and would create an incident reporting system for “catastrophic risk.” AI safety teams and higher education policymakers will care because the bills influence how future AI compliance and disclosures may be measured at state level—especially as schools adopt or restrict AI tools and as regulated procurement expands across public institutions.