What specific challenges arise from AI moderation in Tamil?

Question

Accepted Answer

AI Moderation Faces Significant Challenges with Tamil Language Nuances

Artificial intelligence (AI) systems are proving insufficient for the comprehensive moderation of online content, particularly when dealing with the complexities and nuances of the Tamil language. This limitation is especially pronounced in addressing cyberbullying, where AI may fail to accurately identify harmful or abusive content within the Tamil-speaking community. The Malaysian Communications and Multimedia Commission (MCMC) has highlighted this concern, emphasizing that the reliance on AI alone for content moderation is problematic.
Difficulty in Identifying Harmful Tamil Content

AI's inability to fully grasp the subtleties of certain sentences makes it a less effective tool for detecting cyberbullying and other harmful expressions within the Tamil language. This limitation was specifically communicated to TikTok, which has reportedly shifted towards greater reliance on AI by reducing its human moderator workforce. The minister indicated he would be monitoring TikTok's approach to this issue.
Government Initiatives to Address Online Harm

In response to these challenges and a broader spectrum of online crimes, the Malaysian government has introduced the Child Protection Code (CPC) and Risk Mitigation Code (RMC). These codes have been communicated to social media operators with the aim of preventing platforms from becoming conduits for scams, cyberbullying, and other offenses. The MCMC is actively engaging with platforms, including YouTube in Singapore, to ensure the implementation of these codes.
Penalties and Age Verification Measures

The implementation of age verification through electronic Know Your Customer (e-KYC) mechanisms is underway, and social media platforms that fail to comply with the CPC and RMC once fully enforced will face substantial penalties. The government's stance also includes a directive against children under 16 independently opening social media accounts, though parental creation for educational purposes is permissible. This approach is underscored by the tragic case of influencer Rajeswary Appahu, who died on July 5, 2024, after reporting online threats and defamation.

Reading history

Hide →

This history is cleared when the page is closed.

What specific challenges arise from AI moderation in Tamil?

Loading

Was this article helpful?

Helpful

Not helpful

Report an issue

2

0

0

Share with friends