Singapore's Online Safety Commission (OSC) has been open for business for two months, and it is already busy. Law Minister Edwin Tong told the SHECARES Centre Open House on July 24 that the Commissioner had "a very busy start with about 200 cases already" since the OSC began operations on June 29, 2026, under the Online Safety (Relief and Accountability) Act 2025 (OSRAA) (mlaw.gov.sg).
Two cases Tong described illustrate the model at its best. In one, a survivor found that an intimate recording had been posted without consent to a pornographic website. The OSC issued a takedown direction, and the site complied, removing the video. In another, two minors had their photographs and phone numbers doxxed online; the OSC worked with the platform to remove the posts and suspend the accounts responsible, prompting the children's parents to thank the agency for the speed of the response (mlaw.gov.sg).
What the OSC Actually Does
The OSC, administratively supported by the Infocomm Media Development Authority, is not a court. It is a fast-acting administrative regulator that can issue binding directions to content posters, group administrators, and hosting platforms to disable access to harmful material or restrict accounts — with non-compliance a criminal offense (osc.gov.sg). Its first phase covers five categories: online harassment (including sexual harassment), doxxing, online stalking, intimate image abuse, and image-based child abuse, with eight more categories to be phased in later. A parallel statutory-torts track lets victims sue communicators, administrators, and platforms directly in civil court, with minimum damages of S$5,000 per image or recording for intimate and child-abuse image cases.
Speed is the point. Six "prescribed platforms" — the largest services with significant reach in Singapore — face response deadlines as short as six hours for the most graphic child-abuse and intimate-image content, 24 hours for other intimate-abuse material, and 48 hours for everything else (osc.gov.sg; Baker McKenzie). A separate 2026 SHE survey of 1,048 Singaporeans aged 15 and up found that roughly 40% had experienced some form of online harm in their lifetime, about 25% within the past year, and roughly 40% of those affected reported severe mental, physical, or financial impact (AsiaOne). Against that baseline, a 200-case first month is not surprising — it suggests genuine latent demand for a redress channel that previously ran through slow civil suits or platform customer-service queues.
The Case for the OSC
The strongest argument for this model is that the previous status quo failed victims precisely because it was slow. A parent whose child has been doxxed, or a survivor whose intimate images are circulating, does not have months to wait for a civil suit to work through the courts while the material keeps spreading. Every day a nonconsensual video stays up is a day of compounding harm, and platforms — even well-intentioned ones — often deprioritize individual takedown requests against the volume of their moderation queues. An administrative body with statutory teeth and a six-hour clock for the worst content closes that gap in a way ordinary litigation cannot. Minister Tong's own framing during the bill's second reading was explicit about this trade-off: when MP He Ting Ru proposed additional High Court appeal rights, Tong argued that would "make the process slower, with less finality," discouraging victims from using the system at all (mlaw.gov.sg). That is a defensible institutional design choice, not an oversight.
Where the Design Gets Riskier
The two illustrative cases Tong chose to publicize, however, are the easy ones. Doxxing children and nonconsensual intimate imagery are about as close to unanimous consensus as content moderation gets — nobody credibly argues those posts should stay up, and the harm is unambiguous and irreversible if delayed. The regime's real test is the categories the OSC hasn't yet showcased: harassment and stalking claims that turn on context, intent, and proportionality rather than a clearly identifiable image or a minor's phone number. OSRAA covers 13 categories of harmful activity, several without direct equivalents in comparable UK or Australian online-safety law — a scope choice that reflects, as legal commentary has noted, Singapore's more conservative baseline for acceptable speech.
Minister Tong has argued the safeguards are adequate: platform duties only bite "when an actor fails to act reasonably after receiving proper notice," not through constant pre-emptive surveillance, and courts apply a contextual reasonableness standard rather than "fixed or rigid formulas" (mlaw.gov.sg). That is a reasonable design intent. But it depends entirely on execution once the OSC starts adjudicating harassment complaints that sit closer to public criticism, satire, or unwelcome-but-lawful commentary than to doxxing a child. A six-hour compliance clock built for exposed intimate images is a very different instrument when applied to a contested harassment claim — platforms facing criminal liability for non-compliance have every incentive to comply first and litigate never, which is precisely the over-removal dynamic that has dogged similar takedown regimes elsewhere.
The Right Question Going Forward
Singapore built genuine due-process architecture into OSRAA — administrative review, judicial review avenues, and a Commissioner answerable to Parliament — and the two-month track record on unambiguous harms is a legitimate early success worth crediting. The regulation-skeptic's job now isn't to relitigate whether victims of doxxing and image abuse deserved a faster remedy; they clearly did. It's to watch what happens to the caseload mix as the OSC's remaining eight harm categories phase in, and whether the 200-cases-a-month pace holds up once the docket shifts from unanimous cases to contested speech. That is the data point worth demanding transparency on — a category breakdown of the caseload, not just a running total — before anyone declares the model proven.