Overview
- Anthropic publicly opposed Senate Bill 3444 and is lobbying Illinois lawmakers to change or defeat it, while OpenAI backs the measure as a path to consistent state rules.
- SB 3444 would shield developers of "frontier" AI models from lawsuits over critical harms, defined as events like 100 or more deaths or at least $1 billion in property damage, if the company posts a safety protocol and transparency report and did not act intentionally or recklessly.
- The bill lets a developer count as compliant by adopting European Union safety requirements or by signing a qualifying agreement with a U.S. federal agency, and it would stop applying if overlapping federal law takes effect.
- Anthropic testified in favor of a competing bill, SB 3261, which would require public safety and child‑protection plans for frontier models and independent third‑party audits to test those plans.
- Governor JB Pritzker’s office signaled opposition to giving big tech a full shield from responsibility, and policy analysts say SB 3444 faces long odds, a debate that could shape whether victims can sue after a large‑scale AI harm or whether firms face audit‑based oversight.