Skip to content
GCC AI Research

Roblox expands child safety controls, but Middle East chat remains off as Arabic AI catches up - The National

The National · · Notable

Arabic AI NLP Product Ethics Policy

Summary

Roblox is expanding its child safety controls globally, but live chat functionality remains disabled for users in the Middle East. This decision is due to the current limitations of AI technology in accurately and safely moderating complex Arabic user-generated content. The article indicates that while Arabic AI development is progressing, it has not yet reached the maturity required for real-time, nuanced content moderation critical for child safety. Why it matters: This situation highlights the significant gap in advanced Arabic natural language processing capabilities needed for high-stakes content moderation, affecting product feature availability and user safety in the region.

Keywords

Roblox · child safety · content moderation · Arabic AI · natural language processing

Read original article →

Get the weekly digest

Top AI stories from the GCC region, every week.

Related

Culturally Aware GenAI Risks for Youth: Perspectives from Youth, Parents, and Teachers in a Non-Western Context

arXiv ·

A study investigated the culturally aware risks of Generative AI for youth aged 7-17 in Saudi Arabia, focusing on privacy and safety challenges. Researchers analyzed 736 Reddit posts, 1,262 X (Twitter) posts, and conducted interviews with 31 Saudi participants including youth, parents, and teachers. Findings highlighted context-dependent risks, particularly regarding the disclosure of personal and family information that conflicts with culturally rooted expectations of modesty, privacy, and honor. The study proposes design implications for inclusive, context-sensitive parental controls that align with local cultural norms and values. Why it matters: This research is crucial for developing AI tools and policies that are culturally appropriate and safeguard youth in non-Western contexts like the Middle East.

Polyglot programs: NLP for Arabic and the globe’s diverse dialects

MBZUAI ·

MBZUAI researchers presented studies at EMNLP and ArabicNLP conferences on improving NLP for diverse languages, especially Arabic. One study evaluated ChatGPT and GPT-4's performance across Arabic dialects, finding limitations compared to English. GPT-4 showed better performance than GPT-3.5 in Arabic. Why it matters: This research highlights the need for NLP models to better support the linguistic diversity of Arabic and other languages to avoid widening existing technological gaps.