The Center for AI and Digital Policy, a tech ethics group, has filed a complaint against OpenAI with the FTC.
The center is urging the federal agency to investigate OpenAI and force it to pause any further commercial releases of its large language models.
- The center, led by president Marc Rotenberg, referred to OpenAI's GPT-4 as "biased, deceptive, and a risk to privacy and public safety."
- In its complaint, the group argued that GPT-4 is capable of generating propaganda and malicious code. It cited the likelihood of stereotypes being baked into the LLM's training data, resulting in biases in areas like hiring.
- The ethics center is accusing OpenAI of violating the FTC's AI product guidance as well as Section 5 of the FTC Act, which outlaws unfair and deceptive business practices.
- Rotenberg told Bloomberg that a "practical solution" for unbridled AI advancement is required, "and that practical solution comes from the Federal Trade Commission."
- He said OpenAI needs to be enjoined to halt further GPT releases until there are "adequate safeguards."
- OpenAI did not immediately respond to media requests for comment.
- Yesterday, its CEO Sam Altman tweeted what he believes are the things needed "for a good AGI (artificial general intelligence) future."
- These included coordination among prominent AGI efforts and a global framework for AI regulations, "including democratic governance."
