20 Countries Cap Social Media for Kids, Indonesia Moves First

Seven months after the child-protection rules took effect, Indonesia's Ministry of Communication and Digital counted 28 million child accounts as now protected. On 6 September 2026 the government had also finished verifying 60 products, services and features rated high risk. That first number sounds large. The scope is larger still: at least 20 countries now have rules limiting children's age on social media, and each has picked a different threshold.

The distance between "has a rule" and "the rule is in force" is the part that often goes missing from news coverage. Most of those 20 countries have legislation on paper only, and several only passed it in the past year. There is also a widely circulated claim in Indonesia that needs correcting first.

What the rules actually say in Indonesia

The legal material is layered, and that layering is what short news reports leave out.

The top layer is Government Regulation No. 17 of 2025 on the Governance of Electronic System Providers in Child Protection, known as PP Tunas. President Prabowo Subianto signed it on 28 March 2025, a full year before enforcement began.

On 6 March 2026 the ministry issued three things at once: Ministerial Regulation No. 9 of 2026, plus Ministerial Decree No. 140 and No. 142 of 2026. Decree 140 designates eight services as high-risk products. Decree 142 sets out the technical guidelines for risk assessment and the verification procedure.

The eight are YouTube, TikTok, Facebook, Instagram, Threads, X, Bigo Live and Roblox. Roblox is in the list despite not being social media in the ordinary sense, which shows that what gets rated risky is how children interact with a service, not what category the product sits in.

Two details of PP Tunas go unmentioned most of the time. First, every social networking and social media service is classified high risk by default; a service only escapes that status if it is reassessed by the platform and by the minister together. Second, the assessment uses seven risk dimensions, ranging from a child's possible contact with strangers and threats to personal data, through addiction risk, to possible psychological or physiological harm.

Penalties for platforms escalate in steps: a written warning, a fine, a temporary suspension, then termination of access. A platform can be summoned up to three times before penalties are applied, and it can be held in absentia.

Indonesia moved first in ASEAN, not in the world

This correction matters, and Tempo fact-checked it on 6 April 2026.

The claim that Indonesia is "the first country to ban social media for children" has circulated widely on Facebook, TikTok and YouTube since late March 2026. Part of it is right and part is not. Indonesia is indeed the first country in ASEAN to impose this restriction.

But Indonesia is not a global pioneer. Australia came first: its law was passed in 2024 and took full effect in December 2025, closing the door for under-16s on platforms including TikTok, YouTube, Instagram and Facebook. The Indian state of Karnataka also bans access below 16. An AFP tally dated 21 July 2026 shows five countries with restrictions genuinely in force, while most of the rest are still at the draft bill stage.

The difference is not about severity. Australia has a single federal law that applies at once. ASEAN has a patchwork assembled country by country, with different schedules and different age thresholds.

The global map: what is actually in force

Count the table yourself and the arithmetic is uncomfortable: of nineteen rows, five are genuinely in force. The other fourteen are still drafting a bill, or have not settled on a number at all. A headline claiming "20 countries restrict children's social media" counts both as one thing, when in reality only a quarter of them protect children today.

Why the age thresholds differ

The thresholds of 13, 14, 15 and 16 come out of three different mechanisms, and the mechanism chosen decides who actually gets locked out.

The absolute ban model locks an entire age band with no exceptions. France and the EU lean this way, with the EU proposing a full ban below 13 and tiered access from 13 to 18.

The parental consent model lets a child use a platform as long as there is permission. Germany uses this rule: children aged 13 to 16 may only use social media with parental consent. Italy has its own variant at 14.

The deferred access model is the one most of Southeast Asia uses, and it does not cut children off the internet. It only delays access to high-risk services until a given age, and permits lower-risk services from 13. Indonesia uses that model, and Communication and Digital Minister Meutya Hafid stressed that the policy does not restrict children's internet access as a whole, only delaying access to high-risk platforms.

Enforcing a rule after it is passed

A good law does not become a working law on its own. In Indonesia the early evidence is about leakage.

On 9 April 2026 the ministry sent a warning letter to Google because it had not shown any intention of restricting under-16 access on YouTube. Meta, by contrast, was treated as compliant after amending its community guidelines, which brought the minimum age for Instagram, Facebook and Threads to 16. Eight social media companies operating in Indonesia were reported as compliant.

At the same time, parents interviewed by ABC Indonesia said their children could still open Roblox despite the ban. The ministry itself acknowledged that implementation was still in a transition phase, citing Article 49 of PP Tunas, which allows a transition period of up to two years from promulgation.

The most obvious gap is evasion. Getting past an age limit is as easy as installing a VPN, and no age verification system is genuinely proof against that. The two-year transition does give platforms room to build the mechanisms, but it does nothing to close the easiest shortcut.

What this means for parents in Indonesia

The official justification rests on figures that are not small. UNICEF data cited by the government puts roughly 50 percent of children active online as having been exposed to sexual content, and 42 percent reporting that being online left them feeling scared or uncomfortable. Government records also note 1.45 million cases of online child exploitation.

On the other side, Amnesty International Indonesia has criticised the policy as potentially taking away tens of millions of young people's ability to communicate, access information, develop creatively and express themselves.

Neither figure cancels the other. Both are true, and it is the tension between them that makes the policy hard. What is being protected is access to harmful content; what is being preserved is a child's need to be present in digital space at all. A total ban that is not matched by adequate digital literacy will never resolve it.

Analysis: three things rarely discussed

None of the three points below appears in news coverage, and all three are more useful than repeating the country list.

• The thresholds are not uniform, and that is by design. The range from 13 to 16 exists because each country reassesses risk differently, not because there is a global consensus number. Quoting a single "children's age limit worldwide" is always wrong.
• What gets rated is a product, not a company. PP Tunas assigns status per product, service and feature, with 60 items verified as of 6 September 2026. The consequence is that one company can simultaneously run a low-risk service and a high-risk one. YouTube is on the list not because all of it is dangerous, but because some feature inside it was judged unsafe for children.
• Volume is moving fast and verification quality is not yet public. Twenty-eight million accounts in seven months sounds impressive, but no public figure says how many were genuinely deactivated and how many were merely hidden. Without that, it is impossible to tell whether the policy reduced exposure or only pushed children somewhere else.

The objection most often raised, that bans are imposed without giving parents the means to enforce them, touches the same problem from the other side. The issue is not the direction of the policy but whether the groundwork was done.

The first six months of PP Tunas are best treated as data rather than as a conclusion. The 28 million account figure shows the intended scale. The 1.45 million exploitation cases show an unresolved problem. Both are true in different senses, and only the second one determines whether this policy deserves to be extended.