Attorneys General Warn Openai ‘harm To Children Will Not Be Tolerated’

California Attorney General Rob Bonta and Delaware Attorney General Kathy Jennings met pinch and sent an open letter to OpenAI to definitive their concerns complete nan information of ChatGPT, peculiarly for children and teens. 

The informing comes a week aft Bonta and 44 different attorneys wide sent a missive to 12 of nan apical AI companies, pursuing reports of sexually inappropriate interactions betwixt AI chatbots and children. 

“Since nan issuance of that letter, we learned of nan heartbreaking decease by termination of 1 young Californian aft he had prolonged interactions pinch an OpenAI chatbot, arsenic good arsenic a likewise disturbing murder-suicide successful Connecticut,” Bonta and Jennings write. “Whatever safeguards were successful spot did not work.”

The 2 authorities officials are presently investigating OpenAI’s projected restructuring into a for-profit entity to guarantee that nan ngo of nan nonprofit remains intact. That ngo “includes ensuring that artificial intelligence is deployed safely” and building artificial wide intelligence (AGI) to use each humanity, “including children,” per nan letter. 

“Before we get to benefiting, we request to guarantee that capable information measures are successful spot to not harm,” nan missive continues. “It is our shared position that OpenAI and nan manufacture astatine ample are not wherever they request to beryllium successful ensuring information successful AI products’ improvement and deployment. As Attorneys General, nationalist information is 1 of our halfway missions. As we proceed our speech related to OpenAI’s recapitalization plan, we must activity to accelerate and amplify information arsenic a governing unit successful nan early of this powerful technology.”

Bonta and Jennings person asked for much accusation astir OpenAI’s existent information precautions and governance, and said they expect nan institution to return contiguous remedial measures wherever appropriate.

TechCrunch has reached retired to OpenAI for comment.