A S$2 Million-a-Day Problem
Singapore's Parliament passed the Scams (Countermeasures) and Other Matters Bill on September 9, 2026, and the number that framed the debate is stark: Singaporeans lost S$410.6 million to scams in the first half of 2026 alone — about S$2.3 million a day, even after a 17.9% year-on-year improvement from H1 2025's S$500.2 million (Fintech Singapore, citing Singapore Police Force data). Online platforms were the first point of contact in 89% of reported cases, and Facebook, WhatsApp and Instagram alone accounted for 34.1% of all scams. Against that backdrop, a government moving to arm police with faster, harder enforcement tools is not an overreaction — it's a proportionate response to a persistent, quantified harm.
What the Bill Actually Does
The legislation amends the Protection from Scams Act and the Online Criminal Harms Act (OCHA) in three main ways, per the Ministry of Home Affairs' summary:
- Account Disabling Orders (ADOs). Police can now order service providers to disable a suspicious account for up to 30 days, extendable once, before any crime has been completed — a preemptive, not merely reactive, power.
- Automated takedown directions. OCHA enforcement can now run through computer programs rather than manual review for each case, letting authorities act "at significantly greater speed and scale" against fake ads, accounts and sites.
- Steeper penalties. Maximum fines for designated platforms that fail to comply with OCHA Codes of Practice rise from S$1 million to S$10 million per violation, with daily penalties up to S$300,000 for continuing non-compliance (MHA; Identity Week).
A new offence targets "account mules" — people who sell or hand over personal credentials, including Singpass logins, to let scammers open accounts — with jail terms of up to three years. Police also gain Service Limitation Orders, letting them lock repeat offenders out of self-service banking, telecom and digital-ID channels for up to three years, and a National Scams List, built with defence-tech agency HTX, to pool scam data across banks, telcos and platforms.
The Platforms Already Named
This isn't a hypothetical framework — it lands on top of Codes of Practice the Singapore Police Force issued on August 17, 2026, under existing OCHA authority. A new Online Messaging and Conferencing Services Code covers WhatsApp, Telegram, WeChat, iMessage, FaceTime, Google Messages and Meet; a new Social Media Services Code covers Facebook, Instagram and TikTok; and an enhanced E-Commerce Services Code covers Carousell, Facebook Marketplace and Facebook Business Pages (SPF). Compliance is due by January 31, 2027 (anti-spoofing measures by September 30, 2026). The Bill's fine increase is the enforcement teeth behind codes that were already written; TikTok was added to the designated list in September 2025 after a documented 240% jump in scam cases on the platform.
The Case Against Reflexive Alarm
A regulator moving to pre-emptively disable accounts and automate takedowns is the kind of power that, described abstractly, sounds like a template for censorship. It's worth taking that concern seriously rather than dismissing it: automated enforcement at scale inevitably produces false positives, and a frozen account — even temporarily — can cut someone off from wages, rent payments or emergency funds. Singapore's own Parliament pressed exactly this point during the two-day debate. Senior Minister of State Goh Pei Ming was asked directly what happens to people wrongly caught by an ADO or Service Limitation Order, and offered specific commitments: individuals under a Service Limitation Order retain in-person bank branch access even when self-service channels are cut, salaries and government payouts continue unaffected, and those with disabled accounts can apply to police for a "reasonable amount" of emergency funds (AsiaOne). That's a real, if narrow, safeguard — the harm is bounded to inconvenience rather than destitution, and it survived unanimous passage rather than being brushed aside.
Where the Bet Still Needs Watching
The unresolved question is speed symmetry. An algorithm can disable an account in milliseconds; Goh's pledge that appeals will be handled "fairly and as quickly as possible" is not yet backed by a published SLA, an independent reviewer, or a published error rate for the automated takedown system once it's live. A S$10 million fine plus S$300,000-a-day penalties will predictably push designated platforms toward over-compliance — errs-on-the-side-of-takedown moderation of borderline ads and accounts, since the cost of under-enforcement now dwarfs the cost of a wrongful block. That's a rational platform response to the incentive structure Parliament just built, and it's the actual place where legitimate speech and commerce risk getting caught in the net — not in the mule-jailing provisions, which target conduct (selling access to an account) rather than expression.
The Right Standard
People of Internet's view: the core bet here — faster, harder enforcement against a documented S$800M+/year fraud problem, paired with concrete emergency-access carve-outs — is a defensible one, and Singapore's Parliament did more due diligence on the mechanics than most legislatures managing comparable platform-fine regimes elsewhere. The test now shifts from legislative text to operational reality. If the appeals infrastructure that Goh promised materializes with published timelines and an independent audit trail before the January 2027 compliance deadline, this becomes a credible model for proportionate scam enforcement. If it doesn't — if "as quickly as possible" turns out to mean weeks — the automation that makes enforcement fast will be the same automation that makes wrongful takedowns hard to unwind.