The Office of the Privacy Commissioner for Personal Data, recently completed a review of 60 institutions to understand the application of artificial intelligence. Among them, 95% of the institutions use artificial intelligence in their daily operations. Privacy Commissioner for Personal Data, Ms Ada Chung Lai-ling, said in a radio program today (May 20) that this is a 15-percentage-point increase compared to the compliance review results in 2025, reflecting the growing popularity of artificial intelligence in various industries.
Chung said that nearly 80% of the institutions have been using artificial intelligence for more than a year, indicating that artificial intelligence has gradually become an important part of business operations. Among them, about half of the institutions use three or more artificial intelligence systems, such as answering questions in customer service and creating images and videos in marketing.
Chung pointed out that among the institutions using artificial intelligence, over 40% collect or use personal data through artificial intelligence systems. All these institutions have provided a Personal Information Collection Statement to the data subjects. Nearly 30% of the institutions will retain the personal data collected in the artificial intelligence system, which has decreased compared to the previous review. She described this as a positive phenomenon, reflecting an increase in the institutions' compliance awareness, as retaining the data is riskier than not retaining it.
Chung stated that in this review, institutions adopted more cautious measures when supervising the operation of artificial intelligence. Approximately 80% of the institutions adopted the "Human-in-the-loop" approach, which means retaining human control during the decision-making process of artificial intelligence, and not allowing artificial intelligence to make all decisions, in order to reduce the possibility of errors or improper decisions made by artificial intelligence.
Related News:
Deepline | 'This has no soul': How Internet mistakes a Monet for 'AIGC junk'
ChatGPT encouraged FSU shooter, victim's family alleges in new lawsuit
Comment