There must be a point to age checks online

The Straits Times, There must be a point to age checks online

By Associate Prof Roy Ka-Wei Lee, Information Systems Technology and Design (ISTD)

We need to build safety into digital products, include young people in their design, and encourage platforms to compete on trust.

A 14-year-old opens a social media app after school. The platform verifies the age and automatically places the teenager’s account in a private setting – a process known as “age assurance”.

That is a useful safeguard. On Instagram, for example, a teen account can, by default, limit who may send messages, restrict sensitive content, filter offensive words, remind the user to take breaks and mute notifications overnight. These settings show what age-appropriate design can look like. But age assurance does not, on its own, ensure that every aspect of the user experience has been designed with young people’s safety in mind.

Platforms must still consider who may contact younger users, how notifications should operate, whether reporting tools are understandable, and how recommendation systems should respond to their age and circumstances.

Recent cases in Singapore, including that of a self-radicalised 14-year-old whose online exposure included violent content surfaced through algorithmic recommendations, underscore why safety must consider not only individual posts, but also the pathways through which young users encounter harmful material. Those pathways can also lead children towards cyberbullying, sexual content, self-harm material or content that reinforces eating disorders.

Identifying a younger user matters, but so does designing the experience that follows.

Putting the responsibility on platforms 

Singapore’s approach is to allow social media platforms that meet child-safety standards to continue offering access to users under 18, while those that do not could be blocked. This contrasts with Australia’s minimum-age regime, which requires designated social media platforms to take reasonable steps to prevent users under 16 from holding accounts.

But a differentiated approach like Singapore’s works only if platform safety can be assessed in practice, not simply accepted at face value. The Infocomm Media Development Authority’s (IMDA) Online Safety Assessment Report 2025 shows why this matters: Singapore-linked child sexual exploitation and abuse material cases on X rose from 33 in 2024 to 73 in 2025, while 17 cases of terrorism content were identified on TikTok from Singapore-based accounts.

Minister for Digital Development and Information Josephine Teo has stressed that restrictions are neither the starting point nor the preferred outcome. The priority is to work with platforms to create safer digital environments, she told a recent forum on online safety for children jointly organised by her ministry and the Institute of Policy Studies (IPS).

This move places responsibility not only on young users and their parents, but also on platforms to improve the environments they design. It shifts the policy question from simply whether young people should be allowed onto social media to whether platforms can provide them with an appropriately safe experience.

But age assurance should be only the beginning of a broader approach which puts user safety and rights at the centre of how digital products are conceived, developed and evaluated. That is precisely the problem today: much protection remains a retrofit, introduced after public concern, regulatory pressure or documented harm, rather than being built into products from the start.

Platforms must go further by embedding safety into the design of their products, including recommendation systems, default settings, contact features and reporting mechanisms.

Take Instagram’s Teen Accounts. Safe design is not just verifying that a user is under 18 – it means changing the teen’s actual experience on the platform. This includes making accounts private by default, limiting direct messages to people the teen follows or is already connected to, restricting sensitive content in Explore and Reels, limiting who can tag or mention the teen, filtering offensive words in comments and message requests, prompting the teen to leave the app after 60 minutes, and muting notifications overnight through sleep mode.

Design it safe

Children’s experiences are shaped not only by individual posts, but also by recommendation algorithms, contact settings, notifications, endless scrolling and other features that determine what they see and how long they remain online.

Some platforms have shown that protections can be built directly into products. Meta’s Teen Accounts, for example, place younger users in private accounts and set more restrictive messaging and content settings, while providing time limit reminders and an overnight sleep mode. These measures show safety can be designed into products, but they also remain piecemeal and platform-led rather than a consistent industry standard.

Platforms should assess foreseeable risks before launching or materially changing a service, recommendation system or major feature rather than waiting for harm to occur. Safer defaults, age-appropriate recommendations, limits on contact from unknown adults, understandable reporting channels and meaningful appeals should be part of the product from the outset.

Build it with youth 

Products and policies intended to protect young people should also be designed with them.

The Ministry of Digital Development and Information’s (MDDI) 2025 survey found that users who encountered harmful online content tended not to report it to platforms, with many simply skipping or closing the content. IMDA’s Online Safety Assessment Report 2025 shows why reporting tools still matter: although most platforms improved their response to legitimate user reports, action rates varied widely, and TikTok’s action rate declined from 39 per cent in 2024 to 25 per cent in 2025.

These figures point to a design problem, not just a user-behaviour problem. Consider a teenager who is being mocked by several classmates through comments, edited images and private messages. A reporting tool designed entirely by adults may ask the user to choose between categories such as harassment, impersonation or abusive content.

These labels may be legally precise, but the teenager may not know which category best describes what is happening, whether several incidents can be reported together, or what will happen after the report is submitted.

Involving young people in testing the tool could reveal these problems early. They might suggest simpler language, the ability to report a pattern of behaviour rather than a single post, clearer explanations of the next steps, and immediate options to block, mute or seek support.

Therefore, young people should not be treated only as potential victims. They are experienced users who understand how platforms work in practice, how online conflicts potentially unfold and why some safety features may go unused.

Platforms should involve diverse groups of youths when identifying emerging risks, testing safety tools and evaluating new features. This must go beyond inviting a few young people to endorse decisions already made. Their input should be early, sustained and capable of changing the final product.

Make safety pay
Platforms must also change their mindset.

Too often, safety is treated mainly as a compliance expense: What is the minimum needed to satisfy the law, avoid a fine or prevent a ban?

A better question is: How can being safer make our product more valuable?

Users expect digital services to protect their privacy, well-being and dignity. At the MDDI–IPS forum, some youth participants said they might leave platforms that did not look after their safety. While this reflects the views of a small group, it highlights a broader issue: Safety can influence the confidence of young users, parents, advertisers and regulators.

For platforms, losing that confidence can carry commercial consequences. X’s UK revenue fell by nearly 60 per cent between 2023 and 2024, with the company attributing the decline mainly to large advertisers reducing their spending over concerns about brand safety, reputation and content moderation. While this example concerns platform safety more broadly, it shows that failures of trust can affect a company’s bottom line.

Today, companies compete aggressively for our attention. They should also recognise safety and trust as part of their product’s value, particularly as their ability to serve younger users may increasingly depend on whether they can demonstrate that their platforms are safe.

Regulation and restrictions will still be necessary where platforms fail to meet minimum child-safety standards. Users may find it difficult to leave a platform where their friends, communities and audiences already are. Safety claims are also difficult to compare. This means consumer choice alone cannot be relied upon to drive change.

Independent audits, transparent reporting and clear safety indicators can help users, parents and advertisers distinguish between platforms that genuinely invest in safety and those that merely comply on paper.

When young people enter a platform and declare or verify their age, a safe experience should follow.

• Roy Ka-Wei Lee is an associate professor at the Information Systems Technology and Design Pillar at Singapore University of Technology and Design.