Digital Platform Policy Highlights - Digest 43

Q4 Policy changes: External regulations and rising public pressure have pushed major platforms to roll out new policies aimed at keeping users engaged, safe, and satisfied.

May 20, 2025

This post is part six of a series highlighting how platforms are responding to external regulations in Q4 2024 through policy updates and product adjustments aimed at enhancing compliance and user trust.

TL; DR→ Here are policy changes to address external regulation:

WhatsApp Business Policy Updates in India: Alcohol, OTC medications, Gambling, etc.

The escalating complexity of compliance infrastructure is never more clear: in India, WhatsApp Businesses cannot promote alcohol whereas over-the-counter medications and real-money gaming are allowed, subject to state-specific restrictions. By attempting to enforce state-specific restrictions (e.g., on real-money gaming), Meta is confronting a fundamental platform design challenge: it is unclear if their user identification system was built to track state-level jurisdictions. Naturally some confusion arose, with real-money gaming companies criticizing the lack of clarity, and Meta promising to enhance transparency.(link)

No more AI-Powered Digital Influencers in WeChat’s Live Streaming

By banning AI-generated avatars from its live-streaming platform, Weixin Channel, Tencent’s WeChat has signaled broader efforts to align with China’s regulatory direction while simultaneously signaling their preference to authenticity over scale The idea is to promote authentic real-time interaction between live streamers and viewers, and avoid potential discriminatory and hallucinatory content by AI technologies. Violators may face algorithmic penalty via reduced exposure or restrictions on their e-commerce functions. The timing here is interesting as it diverges from competitors like JD.com and Alibaba who continue allowing AI avatars. (link)

Temu App no longer available on App Stores in Indonesia

Indonesia's demand for app store gatekeepers to block the Chinese e-commerce app Temu Indonesia is another example of nations leveraging intermediary control points to protect domestic markets. The primary fear by the government is that Temu’s direct factory-to-consumer model could harm millions of small businesses in the country. Nations recognizing that platform governance decisions now function as parallel trade policy instruments is not new (e.g., India blocked numerous Chinese apps earlier). As nations strive for digital sovereignty, it is increasingly getting harder to delineate platform governance decisions from public policy decisions. (link)

Temu Stopped in Vietnam Over Registration Issues

Vietnam’s suspension of Temu for failing to complete required business registration procedures, started in November and continues to date (May). The timing here is interesting as it comes amid growing regional concerns about Chinese e-commerce platforms disrupting local retail ecosystems. Some observers wonder if procedural compliance and bureaucratic frictions are being used as tools for managing cross-border platform economics and competition dynamics. TikTok’s e-commerce, on the other hand, is significantly gaining share in Vietnam. (link)

South Korea Fines Meta for Illegally Collecting User Data

The fine of 21.6 billion won (approximately $15 million) on Meta Platforms by Korea's Personal Information Protection Commission for unlawfully collecting sensitive personal information from around 980,000 Facebook users between July 2018 and March 2022 without explicit content. The data, which included details on users' political views, religious beliefs, and sexual orientation, was shared with approximately 4,000 advertisers without obtaining explicit user consent, violating South Korean privacy laws. South Korea's focus on sensitive categories like political views and sexual orientation represents an even deeper challenge to Meta's core infrastructure than what GDPR potentially posed. I cannot help but wonder if $15 million is a meaningful incentive for systemic change. (link)

LinkedIn Suspends Use of Hong Kong Users' Data for AI Training

LinkedIn's suspension of Hong Kong data collection for AI training, following the concerns raised by the Office of the Privacy Commissioner for Personal Data (PCPD), is another example of a complex jurisdictional challenge platforms face when implementing global AI strategies across fragmented privacy regimes. The watchdog objected to LinkedIn’s default opt-in policy, prompting LinkedIn to pause the practice. When an AI model is trained using someone’s data, it is hard to them to ever “forget” things. I wonder what will happen to existing models if some government retrospectively implements a “right to be forgotten” type of policy. (link)

Social Media Ban may Rise Isolation Fears among subsections of Teens.

Australia's proposed teen social media ban has brought to the fore the fundamental tension between risk-minimization governance and platform accessibility economics for vulnerable populations. Some analysts argue that implementing broad access restrictions aimed at reducing mental health harms may put the statistical minority at a disadvantage. For instance, migrant youth population might rely on social media platforms for cultural connection. I think this concern is a fair one: (almost always) binary regulatory approaches to complex platform policies tend to have unintended consequences. (link)

Research help from Aarav Gupta, John Mai, Simran Joshi, Anantesh Mohapatra and Nicole Wu, (Thanks a ton, folks!)

Thank you for reading Platform Policy Research. If you know someone who like this stuff, please share it with them :)

Platform Policy Research