Singapore deepfake ad verification rules force Meta and TikTok identity checks

This regulatory shift moves platform responsibility from reactive content moderation to proactive financial gatekeeping. By requiring identity verification against government records, Singapore is effectively ending the era of anonymous ad buying for high-risk categories. For the broader streaming and social ecosystem, this signals a transition where platforms are legally liable for the authenticity of the entities they monetize, rather than just the content itself. As AI-generated fraud becomes more sophisticated, other jurisdictions may adopt similar 'know your customer' requirements for digital advertising. Watch for whether Meta and TikTok implement these verification tools globally or restrict them to the Singaporean market to manage operational costs.

Meta and TikTok are not alone in facing regulatory pressure over AI-generated fraudulent advertising. In the United States, Meta removed over 2 million accounts linked to scam operations in the first half of 2025, many of which used AI-generated imagery to impersonate public figures in investment fraud schemes. The company reported that scam-related reports from users had declined by 30% following those enforcement actions, though critics noted the measures remained reactive rather than preventive. TikTok content moderation failures, meanwhile, banned over 14 million fraudulent ads globally in 2024 and expanded its AI detection systems in early 2025 to flag synthetic media used in deceptive campaigns, according to its transparency reporting. The regulatory approach Singapore is taking aligns with a broader international trend toward mandatory advertiser identity verification. The European Union's Digital Services Act, which took full effect in February 2024, requires very large online platforms to maintain publicly searchable repositories of political advertisements and verify the identity of advertisers who purchase political or issue-based ads. Australia's Online Safety Act amendments, proposed in late 2025, would impose penalties of up to AU$50 million on platforms that fail to remove scam advertisements within 24 hours of notification, targeting Meta and Google specifically. Singapore's Code of Practice goes further by requiring verification before ads run rather than after complaints are filed, representing a shift from takedown-based enforcement to pre-publication gatekeeping. The technical challenge of detecting AI-generated deepfake scams at scale remains significant for both Meta and TikTok. A study published by Stanford's Internet Observatory in March 2025 found that AI-generated scam videos impersonating celebrities had increased by 400% year-over-year across major social platforms, with investment fraud accounting for the largest share. Meta's own AI detection systems flagged approximately 98% of removed scam content proactively in Q1 2025, though the company acknowledged that adversarial techniques using voice cloning and face-swapping continue to outpace detection capabilities. TikTok partnered with the Global Anti-Scam Alliance in April 2025 to share threat intelligence and develop cross-platform detection standards, signaling industry recognition that no single platform can address the problem in isolation.