US House Of Representatives Starts Banning Staff From Using ChatGPT

House of Representatives (DPR), United States (US) prohibits the use of ChatGPT except for paid versions. (photo: dock. pexels)

JAKARTA - The United States (US) House of Representatives or House of Representatives (DPR) has just stipulated a ban on using ChatGPT except for the paid version.

MPs and staff who want to use the Artificial Intelligence (AI) based chatbot should choose ChatGPT Plus or its paid version.

In her announcement, Head of Administration for the DPR Catherine L. Szpindor said no other versions of the ChatGPT software or any other Big Language Model (LLM) are allowed for use in the DPR at this time.

Because, said Szpindor, the ChatGPT Plus version has important privacy features needed to protect data in the DPR, for which OpenAI as the creator sets a subscription price of 20 US dollars (IDR 300 thousand) per month.

Szpindor emphasized that the use of ChatGPT Plus cannot be arbitrary. Parliamentarians and staff can only use it for research and evaluation purposes.

"ChatGPT Plus has been provisionally authorized by the US House Administration Committee with terms and conditions," Szpindor explained.

"The House office is authorized to experiment with chabot on how the tool might be useful for congressional operations, but staff should not incorporate it into regular workflows," he added.

They are also prohibited from pasting any unpublished blocks of text into the chatbot. Szpindor said, this was to prevent the chat history from being used by LLM anymore.

"The product should only be used with non-sensitive data. Do not paste any unpublished blocks of text into the chatbot. The product must be used with privacy settings enabled," said Szpindor.

"This setting ensures that your history is not stored and your interactions are not fed back into the big language model," he added.

Launching The Verge, Tuesday, June 27, the US House of Representatives announcement came shortly after several technology giants, including Samsung and Apple, restricted or banned employees from using generative AI tools such as ChatGPT.

The company cites concerns that confidential data may leak via the tool due to some previous privacy blunders by OpenAI. Such as a bug that temporarily opens the chat history of each other's users.