Editors’ Note: This is the second in our third annual end-of-year series examining important trends in data privacy and cybersecurity during the coming year. Our previous entry was on energy and security. Up next: trends in state data privacy enforcement.
Since the General Data Protection Regulation (GDPR) came into effect in May 2018, one of the most common questions for practitioners is what the GDPR means for children. As with many provisions of the GDPR, the text itself says relatively little, and precise guidance for businesses – both those intentionally directing online services to children and those that offer more general services that may be used by children – has not yet clarified all of the ambiguities created by the GDPR.
This is to be expected, because the regulatory build-out of the GDPR – in terms of guidance documents, precedents, and other helpful materials – is not yet at the same stage of interpretive and enforcement maturity as the Children’s Online Privacy Protection Act (COPPA) in the United States, for which the Federal Trade Commission has had years to provide explicit regulatory standards and lengthy guidance, and for which there is an abundance of enforcement precedents. I compare below some of the key concepts under both laws because COPPA, though it differs from the GDPR, is conceptually useful in thinking about how online services can approach GDPR compliance.
What does the GDPR require with respect to children?
Article 8 of the GDPR states that processing a child’s data by an online service offered “directly to a child” without parental consent is not permitted, unless “the child is at least 16 years old.” Where a child is below 16, “processing shall be lawful only if and to the extent that consent is given or authorized by the holder of parental responsibility over the child.” Data controllers are required to “make reasonable efforts to verify […] that consent is given or authorized by the holder of parental responsibility over the child, taking into account available technology.” Recital 38 echoes that children can be expected to be “less aware of the risks, consequences, and safeguards” of using online services and so merit extra care when “us[ing] personal data of children for the purposes of marketing or creating personality or user profiles.”
Many of these terms are familiar to anyone who is also familiar with the analogous COPPA Rule, which requires that “[i]t shall be the obligation of the [online service] operator to provide notice and obtain verifiable parental consent prior to collecting, using, or disclosing personal information from children.” 16 CFR 312.4(a). The difference from COPPA, however, is that under COPPA there are elaborate explanations of when a service is offered to a child, who may consent, and by what methods consent can be obtained.
When is an online service offered “directly to a child”?
Asking whether an online service is offered “directly to a child” under the GDPR is similar to asking under COPPA whether a service is “directed to a child,” but with an important difference. As with COPPA, for a generally-available website, a service may be considered to be offered “directly to a child” when it is “made available to all users without any age restrictions” and where the site may reasonably be understood to target children, taking into account such factors as “site content” and “marketing plans.”
But the GDPR parts ways from COPPA in applying only to online services that are in fact “directly” offered. The UK data protection authority, the Information Commissioner’s Office (ICO) has flatly stated than a service “offered through an intermediary, such as a school,” is not offered “directly” to a child. COPPA does not draw this distinction; rather, it applies to any service used by children, but creates special rules around services used by schools and, importantly, allows a school to stand in for a parent in providing consent for a child to use a service in an educational context.
What are “reasonable efforts” to obtain consent?
The GDPR does not define “reasonable efforts” with the specificity that COPPA does. More importantly, UK ICO guidance suggests that the term “reasonable efforts” does not have a static definition; rather, what is “reasonable” depends on the risk of failing properly to identify the individual giving consent. The UK ICO states that “subscrib[ing] to a band’s e-newsletter” is a much lower risk proposition than allowing a child “to post personal data via an unmonitored chat room,” and that the latter calls for “more stringent means to verify the consent.” Interestingly, the UK ICO acknowledges on this point that “[c]ollecting excessive information” for the purposes of consent “is unlikely to comply with the data protection by design approach in the GDPR.” In other words, a data service must collect just enough information to verify consent in light of the risk, but not so much that attempting to verify itself creates a risk of over-intrusion.
So how does a business thread that needle? The answer is not clear. Some companies are understandably re-purposing the same verification methods they use under COPPA, such as requiring a credit or debit card verification. Others are taking less intrusive approaches, which involve less certainty in terms of verification, but also less potentially “excessive” gathering of information. Ultimately, there is no one-size-fits-all solution under the GDPR.
What does “consent” mean?
Consent under both the GDPR and COPPA specifically means informed consent. Under COPPA, this means that the giver of consent must be provided with a direct notice of the online service’s privacy practices, the contents of which are built out by 16 CFR 312.4(c). The GDPR does not provide this level of specificity, but does elaborate on the requisite conditions for consent-based data processing in Article 7, which itself references Recital 32. Recital 32 calls for “specific, informed, and unambiguous” consent from a data subject, following provision of “clear, concise, and not unnecessarily disruptive” notice of what data is to be collected and how it will be processed.
Thus, here too COPPA provides an imperfect, but usable, guide. The basic principles used to develop privacy policies and direct notices under COPPA can also be used to inform consents obtained for GDPR purposes. Both must focus on the important questions of (1) what information is collected, (2) how such information is used, (3) when such information may be disclosed, (4) how a data subject may access or change such information, and (5) how such information is protected. The GDPR is actually more specific in certain regards (for example, as to (4), the data subject must be given a specific set of rights to change and delete data as those obligations are laid out in the GDPR), but in thinking about GDPR compliance, COPPA compliance is not a bad place to start.
What other key differences exist between GDPR and COPPA?
The most important distinction between the GDPR and COPPA is that COPPA is a self-contained regulatory approach to children’s data (although certain states, such as California, do have additional child-focused protections in particular contexts). Article 8 of the GDPR, by contrast, contains special provisions that are peculiar to children, but children are also covered by every other protection of the GDPR as well as member state-level legislation that governs child privacy. This means that, in thinking about GDPR compliance, Article 8 is only one piece of the puzzle. While COPPA provides some useful guideposts and analogies, offering internet services to children resident in the EU calls for a comprehensive approach to privacy that both takes account of the ways that children differ from other data subjects but also the ways that the GDPR protects all data subjects in common.
 The GDPR permits member states to lower this age to 13.