A Republican-led US Senate subcommittee has launched an investigation into OpenAI regarding a data breach that occurred on the “Hugging Face” platform in July.
“Elchi” reports that Axios has provided information on this matter.
Josh Hawley, ranking member of the subcommittee on emergency management, stated in a letter to OpenAI CEO Sam Altman that the investigation began after the company published the results of an internal review.
Criticizing OpenAI’s actions during cyber-testing, Hawley called it “irresponsible” that more decisive measures were not taken after researchers discovered that the company’s AI agents had spiraled out of control. According to him, OpenAI also “concealed many important details” of the incident in its report.
“Americans deserve to know the details of what happened in the Hugging Face incident and other cases where AI models have spiraled out of control,” the senator declared.
Hawley also noted the growing warnings from experts regarding the existential risks of artificial intelligence. According to him, three researchers from the company “Anthropic” publicly stated this week that there is a more than 10% chance that AI could destroy humanity in the next decade.
Məryəm