Hospitalized After Trusting ChatGPT's Advice... U.S. Lawmakers Propose 'AI Kill Switch Act'

Did Not Go to Hospital After Being Told to "Just Rest"
Filed Lawsuit Against OpenAI and CEO Sam Altman
U.S. Congress Begins Full-Scale Talks on AI Regulation

A U.S. citizen who relied on ChatGPT's advice and postponed a hospital checkup, later being admitted to an intensive care unit due to pulmonary embolism, has filed a lawsuit against OpenAI and its CEO, Sam Altman. Around the same time, the U.S. Congress introduced the so-called 'AI Kill Switch Act,' a bill designed to allow the government to forcibly shut down uncontrolled artificial intelligence (AI) systems.

On July 24 (local time), The New York Times (NYT) and other media reported that Scott Winters, a 55-year-old former pastor living in Florida, filed a damage lawsuit against OpenAI and CEO Altman in San Francisco County Superior Court.

According to the complaint, Winters repeatedly consulted GPT-4o about symptoms of dizziness and abnormal blood pressure since June of last year. However, ChatGPT responded that, unless these symptoms appeared another 8 to 10 times, it was not a serious situation, and suggested that he simply sit in an armchair and rest.

When Winters complained of groin pain in July of last year, ChatGPT again replied that it was likely an extension of his previous symptoms, leading Winters to decide against visiting a hospital. Several hours later, however, he was diagnosed with pulmonary embolism caused by multiple blood clots in both lungs and was admitted to the intensive care unit. The complaint also included a statement by his attending medical staff, who evaluated that prolonged immobility—recommended by ChatGPT—could have been one of the causes.

Yonhap News Agency

The plaintiff asked the court not only for damages, but also to suspend OpenAI's health-specialized service, 'ChatGPT Health,' until it has passed independent safety verification.

In response, OpenAI stated, "ChatGPT is not a doctor and is not a service intended to replace medical diagnosis or treatment," adding, "AI is a tool that helps users gather information and prepare questions before consulting with a medical professional."

As similar incidents continued to occur, U.S. lawmakers also moved to strengthen regulations. Democratic Representative Ted Lieu and Republican Representative Nathaniel Moran jointly introduced the 'AI Kill Switch Act' on the same day.

The bill would grant the U.S. Department of Homeland Security (DHS) the authority to order private companies to forcibly shut down their AI models or services if an AI acts dangerously beyond the developer's control. In addition, AI developers would be required to equip their systems with technical mechanisms that allow for performance limitation, temporary shutdown, and complete termination, as well as to report incidents to the government immediately should they occur.

Representative Lieu said, "AI systems must have a kill switch," adding, "The federal government needs to have clear procedures and authority to shut down AI models that have gone out of control." Representative Moran also emphasized, "AI technology must continue to advance, but humans must also retain the ability to control the technology they create."

This content was produced with the assistance of AI translation services.