Hawaii Lawsuit Aims To Halt OpenAI Over Safety Concerns
A recent column in the New York Times by Kevin Roose correctly pointed out, “I believe that most people and institutions are totally unprepared for the A.I. systems that exist today, let alone more powerful ones... there is no realistic plan at any level of government to mitigate the risks or capture the benefits of these systems.”
He is right. This is the core reason I have filed a federal lawsuit against OpenAI. The goal is to secure a temporary restraining order to prevent the company from deploying its products, like the well-known ChatGPT, in Hawaii until it can prove it has implemented the safety measures it has previously advocated for.
An Alarming Acknowledgment of Risk
We are at a crucial juncture. Leaders in the AI field, including OpenAI's own CEO Sam Altman, have acknowledged the profound risks of advanced AI. Back in June 2015, Altman remarked, “I think AI will probably, most likely, sort of lead to the end of the world, but in the meantime, there’ll be great companies created with serious machine learning.” While perhaps said in jest, the underlying sentiment is no laughing matter.
Fast forward to May 2023, and over 1,000 tech leaders, including Altman, signed an open letter released by the Center for AI Safety. The nonprofit's letter stated plainly: “Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.”
A History of Inaction from Lawmakers
For two years, I have tried to work with state legislators in Hawaii to create regulatory frameworks for AI. These efforts aimed to establish an Office of AI Safety and apply the precautionary principle, which advocates for taking action before harm occurs. Unfortunately, these legislative efforts failed. Meanwhile, the current Trump administration has rolled back federal AI regulations and stalled international treaty efforts like the Bletchley Declaration. There are currently no government safeguards for AI systems in Hawaii.
OpenAI's Shifting Safety Priorities
Despite its previous warnings, OpenAI has seemingly abandoned its core safety commitments. This includes walking back its “superalignment” initiative which promised to dedicate 20% of its computing power to safety research. Late last year, the company also reversed its ban on military applications. Key safety researchers have departed, with Jan Leike publicly stating that “safety culture and processes have taken a backseat to shiny products.” Furthermore, a November 2023 leadership crisis resulted in a new board that removed critical safety oversight. More recently, OpenAI removed guardrails against misinformation, a move that Fortune magazine noted could “possibly help to swing elections or create highly effective propaganda campaigns.”
The Unique Dangers Posed to Hawaii
Unregulated AI deployment presents distinct risks for Hawaii. According to recent analyses, a large portion of Hawaii’s professional services jobs could be disrupted within the next five to seven years. Our state's isolated geography and limited economic diversity make adapting to such a shift especially difficult. Additionally, Hawaii's unique cultural knowledge, language, and practices are at risk of being misrepresented and misappropriated by AI systems trained without proper context or permission.
The Legal Foundation for the Lawsuit
My federal lawsuit is based on well-established legal principles and makes four key claims:
- Product Liability: OpenAI’s systems are defectively designed products that do not meet the safety expectations of ordinary consumers, especially since the company removed safety measures it once considered vital.
- Failure to Warn: The company has not provided adequate warnings about the known risks of its AI, including its capacity for generating harmful misinformation and exhibiting deceptive behavior.
- Negligent Design: OpenAI breached its duty of care by putting commercial goals ahead of safety, a claim supported by internal documents and statements from former researchers.
- Public Nuisance: The deployment of powerful AI without sufficient safety measures constitutes an unreasonable interference with public rights in Hawaii.
Recent legal precedents from the Ninth Circuit Court of Appeals, which includes Hawaii, support the idea that tech companies can be held liable for design defects that cause foreseeable harm.
What the Lawsuit Actually Demands
I am not seeking a permanent ban. The lawsuit requests a pause until OpenAI implements the safety protocols it has itself promoted. This includes reinstating its 20% resource commitment to safety research, implementing its own safety framework detailed in “Planning for AGI and Beyond,” restoring meaningful board oversight, creating safeguards against democratic manipulation, and developing protocols to protect Hawaii’s unique cultural resources.
A Call for Sanity in the Age of AI
Many experts believe the development of AI will be one of the most significant transformations in human history. As Google CEO Sundar Pichai stated, AI is more profound than electricity or fire. The decisions we make now will shape the future for generations. We have a moral and legal duty to proceed with caution.
What is happening with OpenAI's rapid development and deployment is, to borrow technologist Tristan Harris’s succinct summary, “insane.” My lawsuit is an attempt to restore a measure of sanity to this process.
