OpenAI has testified in support of proposed Illinois legislation that would limit the circumstances under which artificial intelligence companies can be held liable for harmful outcomes caused by their products, according to testimony provided to state lawmakers.

The ChatGPT maker's backing of the bill comes as the legislation would restrict liability even in cases involving what the bill terms "critical harm," which could include mass casualties or significant financial disasters resulting from AI system failures or misuse.

The legislative proposal represents part of ongoing efforts by states to establish regulatory frameworks for artificial intelligence as the technology becomes more widespread across industries. Illinois joins other states considering various approaches to AI governance and accountability measures.

Meanwhile, concerns about AI's impact on human health and decision-making continue to emerge from researchers and industry observers. Neuroscientist Vivienne Ming has warned about potential cognitive effects from AI dependency, suggesting society may face what she terms a "dementia crisis" if current AI adoption patterns continue unchecked.

Separately, Meta's latest AI model, Muse Spark, has drawn attention for requesting users' raw health data and providing medical advice, raising questions about both privacy protections and the appropriateness of AI systems offering health guidance without proper medical oversight.

The Illinois liability legislation reflects broader industry efforts to establish legal protections as AI systems become more integrated into critical infrastructure and decision-making processes across various sectors.