AU-1
Australia’s covered platforms have had a reasonable-steps duty for under-16 accounts since 10 December 2025.
eSafety — social-media age restrictionsAn account restriction, not a ban on all access to content.Governments are converging on age assurance while diverging on thresholds, service coverage and enforcement.
France adopted an under-15 social-media restriction — then had it struck down by its Constitutional Council before it took effect. The UK has committed to an under-16 model from spring 2027. Greece intends to start its own under-15 regime in January. Ireland wants an EU-level decision focused on under-16s but says it may act domestically if Europe does not.FR-1, UK-1, GR-1, IE-1
These are not versions of one law: they use different ages, cover different services and range from political promise to enforceable obligation. They do, however, point to the same operating reality. Governments increasingly expect platforms to know which users are children, apply rules by age and show regulators how those decisions were made.
France’s under-15 restriction was struck down at constitutional review; Australia’s live regime is already testing adequate platform conduct. The question is whether European law converges before companies have to build for several national versions.
The evidence so far suggests not.
| Jurisdiction | Current position | What is still open |
|---|---|---|
| AustraliaIn force | Under-16 account restriction in force since 10 December 2025; platforms must take reasonable steps to prevent under-age accounts. AU-1, AU-3 | eSafety is investigating whether several large platforms have done enough. |
| FranceStruck down (Article 1) | Both chambers adopted an under-15 restriction on 21 July; the Constitutional Council struck down that restriction (Article 1) on 14 August 2026, before it took effect. FR-1, FR-8, FR-9 | Whether a constitutionally narrower redraft — the President’s stated ambition is spring 2027 — preserves the under-15 threshold. |
| GreeceAnnounced | Government has announced an under-15 restriction intended to apply from 1 January 2027, with platform age verification. GR-1 | Legislation and the final enforcement framework. |
| United KingdomCommitted | Government has committed to an under-16 restriction, with initial regulations planned before the end of 2026 and implementation expected in spring 2027. UK-1, UK-2 | Final service scope, age-assurance requirements and implementation detail. |
| IrelandExploring | Government is exploring restrictions focused on under-16s, prefers an EU-level digital age of majority and reserves domestic action. IE-1, IE-2 | Whether Ireland leads through its EU Council presidency, legislates nationally, or both. |
| European UnionForming | Parliament supports a non-binding minimum age of 16; the Commission has received its expert-panel report and has an age-verification app ready for deployment. EU-1, EU-3, EU-4 | Whether the Commission proposes a common age rule, a risk-based model, or supporting infrastructure without one threshold. |
Policy direction is converging. Legal state is not.
The European Parliament proposed a non-binding minimum age of 16 in November 2025. Australia’s under-16 restriction took effect two weeks later. Ireland, Greece and the UK then set out different national positions between February and June 2026, before France adopted its under-15 text in July. What looked like an Australian experiment is becoming a European policy sequence.EU-1, AU-1, IE-1, GR-1, UK-1, FR-1
France’s two chambers adopted the same compromise on 21 July, which Parliament described as definitively adopted. It would have prohibited under-15s from covered social networks from 1 September, with four more months for existing accounts. But two groups of more than 60 deputies referred it to the Constitutional Council on 23 and 24 JulyFR-1, FR-2, and on 14 August the Council struck down that restriction — Article 1 — in its entirety, before it took effect: a disproportionate restriction on freedom of expression, and inadequate legal safeguards for the privacy implications of mandatory age verification. The political negotiation is over; the restriction does not exist. The rest of the law, not Article 1, was promulgated on 24 August.FR-8, FR-9, FR-11
The European question is narrower than “Brussels blocked the ban,” and is separate from the constitutional one. According to the French Senate’s account, the Commission did not reject the under-15 principle. Its 7 July opinion objected to provisions that would have given Arcom an enforcement role cutting across the Digital Services Act framework, and made further non-binding observations on implementation. The Senate said the associated standstill period ran until 10 August — a separate clock from the constitutional one, with no established causal link between the two.FR-5, FR-6
The censured design would instead have had Arcom report suspected breaches to the competent authorities under EU law. That responded to the Commission’s objection; it is now historical record of struck-down machinery, not evidence of Commission acceptance of anything currently in force.FR-7
France therefore exposes two separate constraints on a national threshold: it cannot design enforcement machinery as though the EU framework does not exist, and it cannot draft a blanket, service-indiscriminate access rule without individualised safeguards and survive constitutional review. The President has tasked the Prime Minister with a redraft and states an ambition to complete it by spring 2027 — an intention, not a bill.FR-12
The under-15 threshold was struck down before it took effect. The redraft timetable and its eventual design have not been chosen.
Open the France analysis →The UK has crossed the political decision line. It intends to lay the first regulations before the end of 2026 and prohibit covered services for under-16s from spring 2027. Its model also plans default overnight and feature restrictions for 16- and 17-year-olds, while excluding private messaging and preserving access to education, news and games.UK-1, UK-2
Greece has chosen a lower threshold. Its government has announced an under-15 restriction from 1 January 2027, with platforms responsible for reliable age verification and re-verification of existing accounts. The legislation and final enforcement design are still pending.GR-1
Ireland is earlier in the process but strategically important. It is exploring under-16 restrictions, wants the digital age of majority decided at EU level and says it will act domestically if necessary. It supports privacy-preserving verification through European work and Ireland’s digital wallet. With online safety a priority for its 2026 Council presidency, Dublin can turn a national concern into a European negotiating issue.IE-1, IE-2
These approaches share an objective, not an operating model. France and Greece use 15; the UK uses 16 and adds controls for older teenagers. France is in constitutional review, Greece and the UK are preparing instruments, and Ireland prefers an EU settlement while retaining a national fallback. That is already enough divergence to rule out one hard-coded European control.
Australia’s regime has been live since 10 December 2025. Covered platforms must take reasonable steps to prevent Australians under 16 from creating or keeping an account. The duty sits with the platform, not the child or parent.AU-1
eSafety said platforms removed access to 4.7 million under-16 accounts by mid-December. That shows action at scale, not unique children kept off the services or proof of success.AU-2
By March, eSafety had significant compliance concerns involving Facebook, Instagram, Snapchat, TikTok and YouTube. It identified repeated age-assurance attempts, weak reporting routes and insufficient measures against new under-age accounts. The regulator is gathering evidence for possible enforcement; it has not found all five platforms in breach.AU-3
The commercial test is not whether every child is kept out, but whether the platform took reasonable steps. Australia’s first enforcement outcomes will help define that standard, backed by a maximum corporate penalty of 150,000 penalty units—A$54.6 million at the cut-off.AU-4
The European Parliament supports a minimum age of 16, with parental consent from 13 to 16, but its November report was non-legislative. The Commission’s special panel delivered its final report on 13 July after considering a common age limit, age-appropriate protections and platform responsibility. The Commission promises proposals; it has not chosen a model.EU-1, EU-4
The practical infrastructure is further ahead. The Commission says its open-source age-verification app is ready for deployment and is designed to prove that a user meets an age threshold without disclosing identity.EU-3
Europe may agree on how a platform checks age before it agrees on the age being checked. The legal routes are diverging while the underlying capability begins to converge.
On 26 August 2026, Meta reached two settlements. The larger is a multistate consent judgment — subject to court approval — with a coalition the New York and Colorado Attorneys General each describe as 47 states plus the District of Columbia, Puerto Rico, American Samoa and the Northern Mariana Islands: 51 attorney-general jurisdictions. Meta will pay that coalition at least $12.1bn, rising to $17.1bn if other major social-media platforms reach comparable settlements of their own.META-1, META-2, META-3
Separately, and announced the same day, Texas reached its own settlement with Meta: over $1bn to the state, funding youth mental-health services, digital-literacy initiatives, after-school programmes and school grants. Texas is not among the 51 jurisdictions named in either lead state’s announcement, and Texas’s own announcement makes no reference to the multistate judgment, names no court, and does not carry the “subject to court approval” qualifier — this brief does not apply that qualifier to the Texas settlement. Meta’s own statement describes the combined effort as an agreement with “a bipartisan group of 52 attorneys general.” No single document states the arithmetic outright, but read together the figures reconcile cleanly: the 51-jurisdiction multistate judgment plus Texas’s separate settlement account for exactly 52. This brief treats that reconciliation as its own editorial assessment, not as an independently verified fact.META-15, META-2
The settlement requires product-design changes for under-18 users of Instagram and Facebook: a combined two-hour daily time limit across both apps, overnight access restrictions and limited notifications during school hours, enforced breaks after sustained use, stronger age-assurance verification, limits on social-comparison features such as visible “like” counts and beauty filters, stronger parental controls, and the option of a non-algorithmic, chronological feed. The restrictions must remain in place for at least five years, with a stricter phase if Snap, TikTok and YouTube also settle on comparable terms — and both the settling states and an independent auditor will assess implementation and effectiveness.META-4, META-5, META-6, META-7, META-8, META-9, META-10, META-11, META-12
The states alleged that Meta designed Instagram with features it knew were harmful to children and teenagers in order to maximise their time on the platform, and that it misled the public about the safety of its services. Meta’s own announcement does not admit or deny those allegations. It frames the agreement as building on its existing safety work and explicitly calls the resulting measures a new industry standard, urging TikTok and YouTube to adopt comparable limits — a call it grounds in the observation that a teen restricted on one app can simply move to another.META-13, META-14
Pursuing legislated age restrictions — in some cases outright account bans for under-15s or under-16s — set through national or EU law.
States secured enforceable product-design constraints and financial exposure through consumer-protection litigation, not legislation.
This does not show that Europe’s account-level restrictions are unnecessary. The two approaches answer different questions, and neither has yet been tested through a full enforcement cycle. What it does give European policymakers is a concrete alternative or complement to a hard age threshold: enforceable design constraints, independent supervision and real financial liability, reached without waiting for a single continent-wide statute.
The development worth tracking is not the dollar figure. It is that child-safety product design has moved, in at least one major jurisdiction, from a voluntary platform setting to an externally enforceable operating constraint — the same direction European policy is reaching for, arrived at by a different route.
Test whether age assurance can support thresholds of 15 and 16, different service definitions, re-verification and a jurisdiction-specific evidence trail.
A single French implementation date or one “EU rule.” France is still under review and the Commission has not chosen its model.
Product and engineering own control flexibility; privacy owns data minimisation; legal and compliance own the jurisdiction-status map; public affairs tracks the institutions still shaping the rules.
Separate the reusable proof-of-age capability from the policy logic deciding what a verified age permits in each market.
The expensive mistake would be waiting for legal uniformity and then discovering that the product can only enforce one threshold in one way.
The Commission standstill cited by the French Senate expires; no resulting Commission position is established yet. FR-6
France’s Constitutional Council strikes down Article 1, the under-15 access restriction, on freedom-of-expression and privacy grounds. FR-8, FR-9
The rest of France’s law is promulgated, without Article 1. FR-11
The Commission response to its expert panel should clarify whether it favours a common threshold or a broader risk-based approach. EU-4
The UK intends to lay its first regulations. UK-1
Greece’s announced implementation date. GR-1
The UK’s intended implementation window. UK-1
The French President’s stated aim to complete a constitutionally robust redraft — not a bill, a referral or a compliance deadline. FR-12
Australia’s first enforcement decisions. AU-3
The fragmentation may be temporary. Ireland wants an EU decision, Greece wants national action harmonised with Europe, the Digital Services Act constrains national enforcement, and the Commission is supplying common verification infrastructure. Those forces could create one practical standard despite different statutory ages.
But the Commission has not chosen its model, Parliament’s preferred age is non-binding and three governments are moving on national timetables. Platforms may have to operate through the divergent phase before convergence arrives.
The Commission proposes a binding common threshold, Ireland secures broad Council support and several states adopt the same verification component.
National courts alter different parts of these regimes, the UK or Greece chooses materially different service definitions, or regulators apply different standards to the same controls.
The policy stages above are drawn from government, parliamentary and regulator records. They support high confidence in what each jurisdiction had announced, adopted or implemented by 29 July, updated for France on 4 September against the Constitutional Council’s own decision text.
Implementation remains less certain. France’s under-15 restriction has been resolved — struck down, not merely pending — but its redraft has not: the President’s spring-2027 timetable is a stated ambition, not a scheduled bill. The final UK and Greek instruments, Ireland’s EU strategy, the Commission’s proposals and Australia’s first enforcement decisions remain open, unaffected by this update. The conclusion about divergence is an assessment of those documented positions, not a forecast that every announced measure will take effect unchanged.
Each reference in the brief resolves to the exact factual claim, its public source and the limit on what that source establishes. Assessments and decision implications remain editorial judgments.
Australia’s covered platforms have had a reasonable-steps duty for under-16 accounts since 10 December 2025.
eSafety — social-media age restrictionsAn account restriction, not a ban on all access to content.eSafety reported 4.7 million under-16 accounts removed or restricted by mid-December 2025.
eSafety — implementation timelineA regulator-reported platform count, not audited unique children.eSafety identified significant compliance concerns at five platforms and was gathering evidence for possible enforcement.
eSafety — compliance statement, 31 March 2026Concerns and investigation are not final breach findings.The maximum corporate penalty is 150,000 penalty units, stated as A$54.6m at the evidence cut-off.
eSafety — civil penaltiesThe currency equivalent is dated because penalty-unit values can change.Both French parliamentary chambers adopted the joint-committee compromise on 21 July 2026.
French Senate — legislative recordDefinitively adopted by Parliament; not yet promulgated or in force.The Constitutional Council received two referrals on 23 and 24 July.
French Senate — legislative recordThe outcome was pending when this claim was first recorded (29 July); the Council ruled 14 August — see FR-8.The adopted text restricts under-15 access to covered online social-network services, with specified exclusions.
French Senate — adopted text, Article 1The text contains service exclusions; it is not a total internet ban.The adopted text states 1 September 2026 commencement and a four-month transition for older accounts.
French Senate — adopted text, Article 1 IIConditional on constitutional review and promulgation, not a dependable live deadline.The Commission’s 7 July opinion objected to several provisions; the Senate did not describe the under-15 principle itself as rejected.
French Senate — account of Commission opinionEvidence for the Senate’s account; not grounds to say Brussels approved or blocked the measure.The Senate said the Commission notification standstill was extended to 10 August 2026.
French Senate — account of Commission processNo legal or Commission outcome is inferred from expiry.The post-adoption design has Arcom report suspected breaches to competent authorities under EU law.
French Senate — legislative recordDescribes the design Article 1 would have used. Article 1 itself was struck down (FR-8); this is historical record of the censured text, not a live enforcement role.The Constitutional Council (decision n° 2026-911 DC, 14 August 2026) declared Article 1 — the under-15 social-media access restriction — contrary to the Constitution in its entirety, before it took effect.
Conseil constitutionnel — decision n° 2026-911 DCArticle 1 only; the Council did not rule on the law’s other provisions (see FR-10).The Council’s stated grounds were (a) a disproportionate restriction on freedom of expression and communication, because the ban applied without regard to an individual minor’s situation or a given service’s specific risk, and (b) inadequate legal safeguards for the privacy implications of the mandatory age-verification requirement.
Conseil constitutionnel — decision n° 2026-911 DCBoth grounds were cited in the decision. Whether either alone would independently have been sufficient to strike down Article 1 was not established by this review — treat them as two stated reasons, not as a confirmed joint-necessity requirement.The Council recognised protecting minors and preventing harm to public order as legitimate aims capable of justifying limits on access — it found Article 1’s specific drafting disproportionate, not the underlying policy goal. It did not rule on the constitutionality of the law’s other provisions.
Conseil constitutionnel — decision n° 2026-911 DCThe other provisions are unexamined by this decision, not affirmatively upheld by it.The rest of the law — not Article 1 — was promulgated as Loi n° 2026-813 du 24 août 2026 (JORF n° 0197, 25 August 2026). Article 1 exists in the published text only as a marginal note recording its own censure.
Légifrance — LOI n° 2026-813 du 24 août 2026No under-15 social-media access restriction exists in force under this law.The President tasked the Prime Minister with preparing a legally robust redraft consistent with the ruling and the European framework, and states an ambition to complete the reform by spring 2027.
Présidence de la République — statement, 14 August 2026A stated presidential ambition, not a bill, a referral or a scheduled date. Spring 2027 is when the President says he wants the reform done, not a compliance deadline.Greece announced an under-15 restriction intended for 1 January 2027 with platform age verification.
Greek Government — announcement, 8 April 2026Announced policy; legislation and final enforcement were pending.The UK committed to an under-16 restriction, with regulations planned in 2026 and implementation intended for spring 2027.
UK Government — commitment, 15 June 2026A government timetable, not a rule already in force.The UK working model preserves specified services and adds default controls for 16- and 17-year-olds.
UK Government — rules fact sheetScope remains subject to the final instruments.Ireland is exploring under-16 restrictions, prefers an EU settlement and reserves domestic action.
Government of Ireland — Digital IrelandNo binding Irish national age restriction had been adopted.Ireland made online safety a priority for its 2026 Council presidency.
Government of Ireland — Digital IrelandAgenda-setting opportunity does not establish Council agreement.The European Parliament proposed a non-binding minimum age of 16, with parental consent from 13 to 16.
European Parliament — report, 26 November 2025A political position, not EU-wide law.The Commission said its open-source, privacy-preserving age-verification app was ready for deployment.
European Commission — app announcement, 15 April 2026Ready for deployment does not mean universal deployment or adoption.The Commission’s special panel delivered its report on 13 July 2026 and the Commission said proposals would follow.
European Commission — special panelNo final European legal model had been selected.Meta reached a multistate settlement, subject to court approval, announced 26 August 2026.
New York Attorney General — settlement announcementSubject to court approval; not yet in force. This qualifier applies to the multistate judgment, not Texas’s separate settlement (see META-15).The multistate consent judgment covers 47 states plus D.C., Puerto Rico, American Samoa and the Northern Mariana Islands — 51 attorney-general jurisdictions, and Texas is not among them. Meta’s own statement separately describes a “bipartisan group of 52 attorneys general”; this brief reconciles that figure as the 51-jurisdiction judgment plus Texas’s separate settlement (META-15).
New York AG (“50 other attorneys general” plus NY; “Texas” does not appear in the release); Colorado AG (full 51-name jurisdiction list, no Texas); Meta (“52 attorneys general,” not broken down)The 51-jurisdiction count and Texas’s absence from it are verified directly. The 51+1=52 reconciliation is this brief’s own editorial assessment — no single document states it outright.Texas announced its own, separate settlement with Meta the same day: over $1bn to the state, plus comparable product-design commitments, resolved independently of the 51-jurisdiction multistate judgment.
Texas Attorney General — settlement announcementTexas’s own release does not reference the multistate judgment, name a court, or use “subject to court approval” — that qualifier is not applied to Texas here.Meta will pay at least $12.1bn to the coalition states, rising to $17.1bn only if other major social-media platforms reach comparable settlements.
New York Attorney General — settlement announcement$17.1bn is conditional, never an unconditional payment.Under-18 users face a combined two-hour daily time limit across Instagram and Facebook.
New York AG; Meta — open letter to TikTok and YouTubeCombined across both apps; messaging is carved out.Under-18 access is restricted overnight and push notifications are limited outside daytime hours.
New York Attorney General — settlement announcementA settlement requirement, not a pre-existing Meta feature.Notifications are limited during school hours.
New York AG; Meta — open letter to TikTok and YouTube—The settlement requires enforced breaks — pause and mindfulness prompts after sustained cumulative use.
New York Attorney General — settlement announcementThe settlement’s own requirement; distinct from Meta’s separate product description.Meta must deploy stronger age-assurance measures to verify the age of young users.
Colorado Attorney General — settlement announcementNo specific verification method is stated in either release.The settlement limits social-comparison features for young users, including visible “like” counts and beauty filters.
Colorado Attorney General — settlement announcement—The settlement requires stronger parental controls and an option for a non-algorithmic, chronological feed.
Colorado AG; New York AG—Implementation and effectiveness will be assessed by both the settling states and an independent auditor.
Colorado Attorney General — settlement announcementAudit frequency and publication of results are not specified.The restrictions must remain in place for at least five years, with a stricter phase if Snap, TikTok and YouTube also settle on comparable terms.
New York AG; Colorado AG—The states alleged Meta designed Instagram with features it knew were harmful to children and teens to maximise their time on the platform, and that it misled the public about platform safety. Meta’s own announcement neither admits nor denies this.
Colorado Attorney General — settlement announcementThe allegation is the states’ claim, not an established fact; Meta does not admit it.Meta frames the agreement as a new industry standard and explicitly calls on TikTok and YouTube to adopt comparable measures.
Meta — open letter to TikTok and YouTube, 26 August 2026Meta’s own characterisation of its position, not an independent assessment.The factual spine uses official parliamentary, government and regulator records. Links open the underlying record.
This is a human-edited worked example of the full analysis brief. Factual claims are drawn from the source record above; interpretations and decision implications are identified through the structure of the brief rather than presented as external facts.