NPR: Google Quietly Drafted AI Chatbot Safety Language Now Copied Across At Least 10 State Bills

NPR reports Google has been directly involved in drafting state-level AI chatbot safety legislation, with at least 10 states — including Hawaii, Iowa, Oregon and Washington — introducing 2026 bills that mirror the language of California's SB 243. The disclosure widens the reverse-federalism playbook already attributed to OpenAI and Anthropic to include Google. Critics quoted call it industry capture of the emerging chatbot liability regime.