What did Malay Mail find when it searched for Malaysian songs?

Question

Accepted Answer

Malay Mail Discovers Over 1,000 Malaysian Artiste Songs in AI Training Datasets

A search conducted by Malay Mail on The Atlantic's AI Watchdog revealed the presence of at least 1,000 songs by Malaysian artistes within large datasets used for training generative Artificial Intelligence (AI) models. These identified songs span a wide spectrum of Malaysian music, encompassing mainstream pop stars, veteran performers, and independent musicians. The discovery raises concerns among Malaysian music publishers regarding the potential unauthorised use of local music in AI development.
Extensive Representation of Malaysian Music Found

The AI Watchdog, a search engine created by The Atlantic, allows users to query massive collections of music data. Malay Mail's investigation using this tool identified a significant number of tracks from Malaysia, indicating a broad inclusion of local artistic output in these AI training datasets. The findings highlight the widespread nature of Malaysian music's inclusion, suggesting that various genres and eras of Malaysian music have been catalogued.
Prominent Artistes Featured in Datasets

Several well-known Malaysian musicians have a notable number of their songs listed in the AI datasets. For instance, Datuk Seri Siti Nurhaliza has approximately 106 songs identified, while the iconic late Tan Sri P. Ramlee has around 134 songs. Contemporary international artiste Yuna has 151 songs listed, and the rock band Butterfingers has 73 songs. The datasets also include music from other notable Malaysian acts such as Carburetor Dung, Fish Leong, Alleycats, Jess Lee, Spooky Wet Dreams, Hujan, Ziana Zain, and Dayang Nurfaizah.
Potential Data Sources and Publisher Concerns

The presence of links to music from platforms like YouTube and Spotify within at least one dataset may explain the extensive range of Malaysian discographies found. However, The Atlantic has clarified that inclusion in a dataset does not definitively confirm its use in training an AI model, nor does absence guarantee non-use. Music publishers in Malaysia, represented by entities like Taja Archive Sdn Bhd, are closely monitoring these developments. They are focused on protecting the rights of songwriters and copyright owners and are prepared to assess appropriate actions if evidence of unauthorised use emerges.

Reading history

Hide →

This history is cleared when the page is closed.

What did Malay Mail find when it searched for Malaysian songs?

Loading

Was this article helpful?

Helpful

Not helpful

Report an issue

2

0

0

Share with friends