Taiwan’s Digital Ministry is seeking NT$40 million (approximately US$1.3 million) this year to purchase data from the private sector to expand AI language databases. At a legislative committee meeting on Wednesday, Deputy Minister Ning Yeh (葉寧) announced that they have drafted the first version of "Taiwan AI Sovereignty Training Corpus" licensing terms to streamline data acquisition processes and reduce negotiation costs with individual entities.
The Legislature’s Education and Culture Committee is currently reviewing multiple versions of the "Taiwan’s AI Basic Act" draft, with over ten proposals submitted by legislators from different parties. Although the National Science and Technology Council's (NSTC) version has completed its preliminary announcement phase, it remains under Cabinet review and has not yet been sent to the Legislature, raising concerns about meeting the May 31 legislative session deadline.
To develop Taiwan's AI sovereignty, the Digital Ministry is taking a two-pronged approach: first, focusing on government data through standardized licensing agreements developed in collaboration with the Intellectual Property Office, and second, targeting private sector content. Deputy Minister Yeh indicated that while they plan to request NT$40 million from the NSTC to purchase private sector data, the initial phase will emphasize voluntary contributions and incentive-based participation.
Several legislators have also raised concerns about AI applications in mental health counseling, therapy, and emotional support. They questioned the accuracy and professional qualifications of AI systems like ChatGPT when providing psychological services, suggesting that the Health Ministry should be included in future discussions about AI regulation and development.