Chat PT scrutiny:
Digital Minister Audrey Tan announced the move after some local governments touted their plans to integrate the applications.
-
Sherry Shan / Staff Writer
The National Science and Technology Council (NSTC) is putting together guidelines on the use of artificial intelligence (AI) in government agencies after several local governments announced plans to use ChatGPT to serve their residents. said Digital Minister Audrey Tang. yesterday.
Tang told lawmakers at a Congressional Transportation Committee meeting about the steps the government is taking to prevent personal data breaches as more people pay their taxes online and do business using ChatGPT. made this statement to
Taoyuan City plans to use ChatGPT in its 1999 municipal government hotline, and Tainan City has developed a supplementary education module using AI apps.
Democratic Progressive Party (DPP) legislator Lee Kunze said many governments and private companies around the world have started using ChatGPT, but countries such as Italy have opted for the technology due to privacy concerns. said to ban the use of
He asked Tan whether Taiwan would ban civil servants from accessing ChatGPT and other generative AI applications to perform government duties.
Tan said Italy lifted its ban on ChatGPT after its developer, OpenAI, implemented the changes requested by Rome.
Tan said the NSTC is working on a guidebook on the use of generative AI applications in government agencies.
“The guidebook will set out two main principles. First, government officials should not submit content generated by generative AI systems without scrutinizing it. Unlike , content created by generative AI systems can come out of nowhere: they can be false statements that appear logical,” she said.
Second, using AI systems only when government departments have no internet connection and instead using local servers under full control of government departments can prevent personal data breaches, he said. Stated.
Regarding the risk of data leaks in companies using AI, Tan said companies should always be careful not to leak sensitive information to the public.
“The problem is that most companies rely too much on AI language models available online because they are convenient. There is no need to access online chatbots and no risk of data leaks,” she said.
The National Institute of Cyber Security is building Taiwan’s AI framework and plans to cooperate with the Institute of Industrial Technology on related projects in the second half of this year, he said.
The NSTC plans to release a “trusted AI dialogue engine” to the public by the end of this year, Tan said, adding that the National Institute of Cybersecurity will be responsible for certifying intermediate versions of the dialogue engine.
“Any organization should use local technology or work with a trusted local cloud provider to avoid uploading all their data to the free public cloud, which would lead to data leakage concerns.” she said.
Comments are moderated. Please keep your comments relevant to the article. Speech containing profanity, obscenity, personal attacks of any kind, or promotion will be removed and users will be banned. The final decision is at the discretion of The Taipei Times.
