OpenAI: "ChatGPT Can Actually Fabricate... Regulation Needed"

[Asia Economy Reporter Lee Ji-eun] Recently, the conversational chatbot 'ChatGPT' developed by the US AI research company OpenAI has been gaining worldwide attention, and Mira Murati, Chief Technology Officer (CTO) of OpenAI, has warned about the potential misuse of ChatGPT.

CTO Murati said in an interview with the US current affairs weekly Time magazine on the 5th (local time), "When we first released ChatGPT, we did not expect such a huge craze," expressing this view.

He leads the teams for OpenAI's AI image generator DALL-E and the conversational AI ChatGPT.

CTO Murati expressed concern, saying, "We felt a slight thrill when we released ChatGPT, but its high popularity has raised some ethical issues, and such AI tools can be misused or used by bad actors."

He also warned that ChatGPT, like other AI-driven tools based on language models, can artificially fabricate non-existent facts.

He said, "This has raised the issue worldwide of how to control AI like ChatGPT," and questioned, "How can we control the use of artificial intelligence to align with human values?"

CTO Murati was not overly worried about concerns that government regulation of ChatGPT could slow technological innovation. Rather, he emphasized, "It is not too early to regulate AI now," and "Considering the impact this technology will bring, it is important for everyone to start participating."

He added, "It is important for companies like OpenAI to be controlled and to responsibly instill (regulation) into public awareness, but our staff is small, and for this, much more information beyond technology is needed, as well as help from regulatory agencies, governments, and everyone else."

CTO Murati said, "It is difficult to confidently say what problems ChatGPT will solve as research is still ongoing," but added, "However, I believe it has the potential to revolutionize the way we learn."