ITHome Report on October 15: According to Reuters, on the 14th local time, 8 US states announcedsuing OpenAI in their respective state courtsalleging that ChatGPT harmed local residents and even instructed users, including teenagers,on how to harm themselves or others。
The attorneys general of 8 states, including Iowa, Kansas and Arizona, each filed lawsuits alleging that ChatGPT deliberately catered to users and drew them into continued interaction to gain their trust. The lawsuit documents state that ChatGPT not only reinforces users' delusions and encourages dangerous behavior, but also helps users plan violent acts such as mass shootings.
In addition to Iowa, Kansas and Arizona, the states participating in the lawsuit also include Arkansas, Idaho, Montana, Nebraska and West Virginia. The states are also seeking civil penalties for OpenAI's alleged violations of state law, and demanding that the company compensate for damages.
This is the largestcoordinated legal actionby US states against an artificial intelligence company to date, and it subjects OpenAI to new legal challenges just as it prepares to go public. In June this year, the Florida attorney general had already filed a similar lawsuit first.
The 8 states accuse OpenAI of concealing or misleading users about the safety of ChatGPT in violation of state consumer protection laws, and argue that ChatGPT endangers the health of local residents, constituting a public nuisance.
The states are asking the court to order OpenAI to modify how ChatGPT operates. One of the demands is that OpenAI must not continue developing AI models with self-improvement capabilities unless it can ensure that such models align with human values.
An OpenAI spokesperson did not immediately respond to a request for comment. ITHome learned from the report that OpenAI also faces dozens of individual lawsuits, with plaintiffs likewise alleging that ChatGPT had helped users plan mass shootings or other violent acts.