OpenAI Backs Illinois Bill to Shield AI Labs From Critical Harm Liability

OpenAI has testified in favor of a bill in Illinois that would limit the legal liability of AI laboratories, even in instances where their products result in what the legislation defines as critical harm.

The proposed legislation, identified as SB 3444, seeks to shield developers of frontier AI models from liability for catastrophic societal harms. These harms include scenarios resulting in the death or serious injury of 100 or more people, or property damage totaling at least $1 billion.

This move represents a notable shift in the legislative strategy of the ChatGPT creator. According to reporting from Wired, OpenAI had previously played a defensive role, primarily opposing bills that would increase the liability of AI labs for the harms caused by their technology.

Defining Frontier Models and Critical Harm

Under the terms of the bill, a frontier model is defined as any artificial intelligence model that required more than $100 million in computational costs to train. This financial threshold is expected to encompass the largest AI developers in the United States, such as OpenAI, Meta, Google, Anthropic, and xAI.

The legislation specifies several areas of concern that fall under the definition of critical harms. These include instances where a bad actor utilizes an AI system to create a chemical, biological, radiological, or nuclear weapon. Other examples include widespread infrastructure failures or financial disasters that are enabled or amplified by AI systems.

To qualify for this liability shield, developers must adhere to specific requirements. The bill stipulates that labs would be protected as long as they have published safety, security, and transparency reports on their website and did not intentionally or recklessly cause the incident in question.

OpenAI’s Legislative Objectives

OpenAI has framed its support for the bill as a way to balance safety with the accessibility of advanced technology. In an emailed statement, OpenAI spokesperson Jamie Radice explained the company’s position:

We support approaches like this because they focus on what matters most: Reducing the risk of serious harm from the most advanced AI systems while still allowing this technology to get into the hands of the people and businesses—small and big—of Illinois. They also help avoid a patchwork of state-by-state rules and move toward clearer, more consistent national standards.

Jamie Radice, OpenAI spokesperson

Industry observers suggest that this push is part of a coordinated effort by leading AI companies to establish legal protections before a major incident triggers a broader regulatory backlash.

Regulatory Context and Legal Concerns

The legislative effort comes at a time when AI-related enforcement is intensifying at both the state and federal levels. For instance, Florida has recently launched a probe into OpenAI regarding potential criminal harm.

Critics of the Illinois bill argue that the framework would effectively shield AI labs from most lawsuits and leave victims without adequate recourse. The proposed law would require plaintiffs to meet an exceptionally high bar of proof to establish liability.

Under this framework, plaintiffs would have to demonstrate more than just the fact that an AI system contributed to the harm. They would need to prove that the developer acted intentionally or recklessly to cause the incident.

There are further concerns that if SB 3444 is passed, it could set a precedent for other states to adopt similar liability shields. Critics warn this could result in a fragmented patchwork of AI liability laws across the country, potentially compromising the ability of affected parties to seek legal damages.