Algorithms Are Neither Good Nor Bad, Says Meta Policy Director

The Astana Times provides news and information from Kazakhstan and around the world.

ASTANA — Protecting teenagers online requires more than privacy settings and messaging restrictions, as ensuring that recommended content is age-appropriate remains a challenge, said Philip Chua, Director of Public Policy, Products (APAC) at Meta, in an exclusive interview with The Astana Times.

Chua said Meta is extending its approach to online safety from protective account settings to the content teenagers encounter, including posts from accounts they follow and recommendations generated by algorithms.

“Fundamentally, algorithms are a way of sorting a lot of content that we have to potentially show a person, and figuring out what to show first. Algorithms are neutral. They are neither inherently good or inherently bad,” he said.

Meta has introduced content settings inspired by movie-rating criteria for ages 13 and older, designed to limit teenagers’ exposure to sensitive material across recommended content and accounts they follow.

The approach is part of a wider discussion over how technology companies and governments can protect younger users without unnecessarily restricting digital access. For Kazakhstan, Meta is recommending stronger protections within existing platforms, alongside age verification systems that would reduce the need for users to share personal information with individual applications.

“What we are recommending to governments, including in Kazakhstan, is this idea of an app store or OS-level [operating system-level] approach to age assurance,” he said.

The proposal would allow applications to receive age-related information through app stores or operating systems rather than requiring families to provide the same personal details separately to multiple platforms.

Can social media safeguards prevent harm?

While Meta has expanded protections for teenage users, their effectiveness also depends on how successfully the company can prevent inappropriate material from appearing in recommendations and other content. Some videos, for example, may comply with platform rules while containing suggestive music, language or themes unsuitable for younger audiences.

When asked how Meta addresses such material, Chua explained that the company combines artificial intelligence and human review to enforce its community standards, while applying additional content restrictions to teenage accounts.

He said these measures are designed to reduce exposure to inappropriate content, although technology cannot guarantee that every potentially harmful experience will be prevented.

“So what we would expect is never 100%. I think technology will keep getting better. But our commitment here is to really avoid any content that A, doesn’t belong on our service, but B, also removes a lot of what you would perhaps say is borderline, potentially sensitive, not something that you want somebody 13 and above to see,” he said.

Age verification raises privacy questions

Ensuring that these protections reach the intended users presents another challenge: determining their ages accurately while limiting the collection of personal information.

Chua described age assurance as an industry-wide issue involving technological limitations and privacy concerns, particularly when young people use numerous applications. Meta uses information provided by users alongside technology that analyzes behavioral patterns to estimate whether an account belongs to someone younger or older than 18.

The accounts people follow, for example, can offer clues about their age. 

“The example that I like to use is if you are following Backstreet Boys on Instagram, you are probably slightly older as compared to if you are following Blackpink,”he said.

Such patterns can help platforms estimate ages without relying exclusively on more intrusive identity-verification methods. However, Chua said the number of applications used by young people raises questions about how widely their personal information should be shared.

“And I think, more importantly, the issue that we are seeing is that insofar as a young person can use up to 40 apps every day, should every app be collecting age and personal information? And what does that mean in terms of how that data can be compromised?” he said.

Under Meta’s proposed approach, age assurance would take place at the app-store or operating-system level, allowing applications to provide age-appropriate experiences without separately collecting the same information. According to Chua, parents could provide their children’s age information when first setting up a device, with app stores and operating systems helping ensure that applications apply suitable protections.

“If there is a way that we can, on a one-stop basis, on a once-off basis, just collect that information at the OS level or the app store level, that’s going to protect privacy and streamline the amount of data that needs to be shared with any app,” he said.

Balancing online protection and digital access

The question of how to protect young users while respecting privacy forms part of a broader policy debate over whether governments should strengthen safeguards within existing platforms or restrict access to social media altogether.

“We have seen that a ban on social media actually does not result in a decrease of time that a young person spends online. We have seen that a ban on social media results in young people finding ways to circumvent the system or finding ways to migrate to other apps that may not be regulated, potentially could be less safe, and generally would not have spent the same kind of investment that big social media companies have spent all these years to protecting young people,” he said.

Australia became a prominent test case when restrictions on social media accounts for users younger than 16 took effect on Dec. 10, 2025. Early findings from the country’s eSafety Commissioner indicate a decline in account ownership among children ages 10-15, although many continued accessing restricted platforms. The long-term effects remain under evaluation.

Chua said restrictions could also leave young people feeling excluded from online communities or encourage them to seek alternative platforms with fewer safety measures.

For Kazakhstan, he recommended building on safeguards already introduced by major technology companies, combining restrictive default settings for teenagers with parental choice rather than adopting blanket prohibitions.

“We need to respect the role of parental choice. Those two ingredients, I think, are very important ingredients in how you can have a balanced regulatory framework so that we are not in a situation where we are denying young people access to digital skills and cutting them off from their digital lives,” he said.

Government restrictions and access to online platforms

Government restrictions on social media access, particularly during politically sensitive periods, present a separate challenge involving regulatory authority, platform compliance and the public’s ability to communicate online.

When asked how Meta approaches such restrictions, Chua said maintaining working relationships with governments is central to the company’s public policy activities.

“Our first principle here is really to collaborate and have a very good working relationship with governments around the world because we believe we share a common interest on preserving access to the internet, preserving the digital conversation and that’s going to be a very important part of our work,” he said.

Chua noted that discussions with governments often concern content that may be inappropriate or illegal under national legislation. Meta seeks to enforce its own policies while complying with applicable laws, including in Kazakhstan, where content regulation remains part of its engagement with policymakers.

He linked these regulatory questions to Kazakhstan’s ambitions in AI and the digital economy, noting that decisions about online access also have implications for the opportunities available to younger generations.

“We also know that Kazakhstan has a very young population and I think the country has a lot of ambitions for the future and so I think this is why we are talking about AI and innovation and investment into the country at the same time we are talking also about how we can ensure a balanced regulatory solution,” he said.

“We want to make sure that we don’t outright restrict access to important digital skills that are critical for the future,” he added.

The post Algorithms Are Neither Good Nor Bad, Says Meta Policy Director appeared first on The Astana Times.