OpenAI has expressed support for a bill in the state of Illinois, known as SB 3444, which seeks to shield artificial intelligence labs from liabilities in cases where their models are used to cause severe social harm, such as mass casualties or large-scale financial losses.
The proposal stipulates that AI developers will not be held accountable for critical damages, provided they have not acted intentionally or negligently and that they publish safety and transparency reports on their official websites.
This stance marks a significant shift in OpenAI’s legislative approach, as the organization previously opposed initiatives that could impose liabilities on labs for damages stemming from their technologies.
According to Wired, SB 3444 could set a precedent for the industry, being bolder than other measures OpenAI has endorsed in the past.
The bill defines a “border model” as any AI system trained with computational costs exceeding $100 million, a criterion that would encompass major U.S. labs, including OpenAI, Google, xAI, Anthropic, and Meta.
Jamie Radice, an OpenAI spokesperson, argued that the initiative aims to prevent the fragmentation of state regulations, promoting more uniform and consistent national standards for the sector.
Although federal legislation on AI developers’ liability has yet to be passed in the U.S., the debate is gaining urgency with the release of increasingly advanced models, which pose new risks related to safety and cybersecurity.
Caitlin Niedermeyer, from OpenAI’s Global Affairs team, advocated for the need for a federal regulatory framework that balances innovation and safety without compromising U.S. competitiveness in global technological development.
However, there is considerable resistance to the chances of SB 3444’s approval. Scott Wiener, an expert linked to AI regulation debates, pointed out that the bill faces significant opposition in Illinois, a state known for adopting stringent stances on technology.
He indicated that a significant portion of the local population, about 90%, opposes the idea of exempting AI companies from liabilities. Illinois has already discussed proposals in the opposite direction, seeking to increase the accountability of AI model developers.
While SB 3444 focuses on high-impact events, such as financial disasters or mass casualties, AI labs also deal with controversies over individual-scale damages.
Recent cases involving AI platforms and severe psychological impacts on users have led to lawsuits against companies in the sector. The absence of unified federal legislation keeps the regulatory landscape fragmented, with states like California and New York exploring their own measures to demand greater transparency from developers, although some initiatives have faced obstacles, such as vetoes on local projects.
The debate on AI regulation in the U.S. remains open, with OpenAI positioning itself as an active voice in the pursuit of rules that protect innovation while addressing the inherent risks of technology. The evolution of SB 3444 in Illinois will be a crucial test for the future of liability in the sector.
Original published at O Cafezinho.