The Federal Trade Commission has opened an investigation into OpenAI, the creator of chatbot ChatGPT and text-to-image generator DALL-E.
The FTC sent a 20-page letter to OpenAI this week in which it sought information from the company about its security practices, the steps it took to modify, supplement, or omit content from its datasets, and how it obtained the data its models were trained on.
The letter was first obtained by The Washington Post and confirmed by The New York Times. The FTC denied a Freedom of Information Act request filed by Artnet News seeking the document.
“We have no comment,” FTC spokesperson Juliana Gruenwald said in an email to Artnet News. “FTC investigations are nonpublic so we generally do not comment on whether we are investigating a particular matter.”
The letter did not specify that the investigation was limited to ChatGPT, a large language model product that uses neural networks to generate text and is reinventing how online search engines function.
Relevant to artists, the company also makes the artificial intelligence image generator DALL-E, first introduced in January 2021. A.I. image generators have swiftly grown in popularity for their ease and accessibility, while sparking ire for their increasingly widespread usage.
DALL-E and other A.I. art generators have also come under fire from artists, who claim that their art was used to train the models without their permission and thus violates their copyrights. In January, a class action lawsuit was launched by a group of artists against the companies behind Midjourney, DreamUp, and Stable Diffusion, alleging these platforms breached copyright and unfair competition laws.
The FTC investigation could pose challenges for OpenAI’s DALL-E product beyond ChatGPT, as the probe is seeking information on “each large language model product” the company offers.
In May, OpenAI’s CEO Sam Altman testified before a Senate Judiciary Committee, where he urged lawmakers to regulate and “mitigate the risks of increasingly powerful models.” He added: “I think if this technology goes wrong, it can go quite wrong. We want to work with the government to prevent that from happening.”
More Trending Stories:
An Israeli First-Grader Stumbled on a 3,500-Year-Old Egyptian Amulet on a School Trip