By Ronald G. London and David Silverman
As 2012 drew to a close, the Federal Trade Commission (FTC) issued amended rules updating its regulations that implement the Children’s Online Privacy Protection Act (i.e., the COPPA Rule). The FTC’s extensive changes to the COPPA Rule are intended to reflect technological developments and evolving popular online practices – primarily, social networking, smartphone Internet access, and use of geolocation information. The Rule changes include amendments to the definitions of “operator,” “personal information,” and “websites or online service directed to children” critical to when the Rule is triggered and how it operates, and updates to requirements for providing notice and getting consent from parents for the collection and use of children’s online personal information, and for maintaining its confidentiality and security, as well as new provisions addressing data retention and deletion.
In short, the Rule amendments, which take effect starting July 1, 2013, seek to “broaden and clarify” the regulations in a variety of contexts. And insofar as COPPA is the primary statute affording the FTC specific regulatory authority over the use of online personally identifiable information, these rule changes could resonate more broadly as reflections of the FTC’s positions on privacy protection generally, and also could force changes in the practices of online services and websites that do not target children directly. As described in this article, the new changes amend the COPPA Rule so that it:
- Covers data collection by plug-ins, software downloads, or advertising networks integrated into websites
- Requires parental consent to include persistent identifiers that can be used for behavioral advertising and other tracking across web sites
- Allows persistent identifiers to be used without parental consent if they are limited to a white-listed set of “internal” operations that include delivery of contextual ads, frequency capping, anti-fraud measures, compliance efforts, and authenticating users or personalizing content, or other “internal” operations specifically approved by the FTC
- Treats geolocation information and any photo-, video- or audio-file that includes a child’s image or voice as personal information
- Allows certain child-directed sites and services to differentiate among users by age-screening, and to provide notice and obtain parental consent only for those self-identifying as under 13
- Streamlines the notice requirements to ensure information about data collection and use practices is presented to parents in a succinct “just-in-time” notice
- Offers more options in the non-exhaustive list of acceptable methods for obtaining prior verifiable parental consent
- Includes new exceptions from the parental notice and consent requirements
- Requires “reasonable steps” to ensure that any release of a child’s personal information is made only to service providers and third parties capable of maintaining its confidentiality, security, and integrity
- Regulates data retention and deletion
- Strengthens FTC oversight of self-regulatory safe harbor programs
- Establishes voluntary a pre-approval process for new consent methods
COPPA Basics and the Need for Updated Rules
COPPA requires notice to parents and verifiable parental consent prior to the online collection of any personal information from children under the age of 13. Because it was first enacted in 1998, and the FTC first adopted implementing regulations in 1999, the rules have been due for updating for some time.
The FTC issued a notice of proposed rulemaking in 2011 covering five different areas. The proposal implicated the Rule’s definitions, including what children’s “personal information” it covers, and what it means to “collect” it. The proposal also set out planned updates to providing parental notice, new parental consent mechanisms, and on how to ensure confidentiality and security of the collected personal information. The proposal further suggested fine-tuning the “safe harbor” for how self-regulatory programs can be deemed “in compliance” with COPPA.
The FTC later issued a supplemental rulemaking notice in 2012 that proposed to treat screen names as personal information, but only when they function in the same manner as online contact information, and to allow the use of some persistent identifiers by expanding and clarifying the types of “internal” operations that do not require notice or consent. The supplemental notice also addressed data collection by plug-ins, software downloads, or advertising networks integrated into website, and “mixed use” sites that attract a disproportionately large percentage of children under age 13.
The amended Final Rule announced by the FTC adopted most of the proposed changes with a few critical departures. Those revisions involve provisions across virtually the entire landscape of the COPPA Rule.
Definition of “Personal Information”
Among the most significant changes are those regarding the definition of “personal information,” which is the focus of COPPA and its implementing rules when it pertains to children and is collected, used, and/or disclosed. Under the former Rule, screen names were considered “personal information” only when coupled with an individual’s email address. As amended, the definition includes any screen or user names that function “in the same manner as online contact information,” for example, as an email address that permits direct contact with the child. The FTC also expanded the “online contact information” component of the “personal information” definition, so that it includes not only email addresses or other “substantially similar identifiers” through which a child may be contacted, but also identifiers for instant messaging (IM), Voice Over Internet Protocol (“VOIP”), and video-chat platforms. The FTC declined, however, to include mobile phone numbers as “personal information” or “online contact information.”
The FTC also added to the Rule’s definition of “personal information” all photographs, as well as video and audio files that contain a child’s image or voice, regardless of whether they contain contact information (though operators who choose to blur images of children prior to their posting can avoid coming within this definitional change). This is a potentially significant addition, as social media, online games, and mobile devices have all made it easier for users to share video and audio files. The new rules also include “geolocation data” that is sufficient to identify a street name and the name of a city or town.
In addition, the FTC expanded the definition of “persistent identifiers” that can be used to identify an individual and thus qualify as “personal information.” The former Rule included as covered persistent identifiers only cookies that are coupled with individually identifiable information. The new rule includes customer numbers in cookies, IP addresses, processor or serial device numbers, and even unique device identifiers. The definition, however, excludes persistent identifiers collected solely to provide “support for the internal operations” of a website or online service, which the FTC adds as a new defined term.
As to that key exception, the Rule’s definition of “internal operations” is expanded – but still limited to – several specified functionalities. These operations now include maintaining or analyzing the functioning of a website or online service, performing network communications, authenticating users or personalizing content for the site or service, serving contextual ads or capping the frequency of ads, protecting the security or integrity of the user, site or service, ensuring legal or regulatory compliance, and fulfilling requests of a child as permitted elsewhere in the rules. To the extent a COPPA-covered site engages service providers to perform these functions, they still qualify under “support for internal operations.” Similarly, if a third party that is deemed an operator or co-operator collects persistent identifiers used to support internal functions, the carve-out still applies. The FTC’s Chairman explained the changes as “closing a loophole” that allowed collection of data from children through plug-ins, noting that while advertisers and ad networks can continue to advertise on sites directed to children, the Rule will limit advertisers and networks by prohibiting the construction of “massive profiles” unless there is parental consent.
In conjunction with the expanded definition of “internal operations,” the FTC created a process for adding to the list of “internal operations” not requiring consent for collection and use of personal information. However, the approach contemplated by the new Rule anticipates detailed filings justifying the additions(s), even if the FTC expedites public comment leading to issuance of a written determination within 120 days of the request’s filing. Moreover, what this means is the specification of functions in the “internal operations” definition is effectively a “white list” of activities is allowed without consent, while all others require consent, unless and until someone files with the FTC, makes a case, and receives FTC approval.
The FTC further revised the persistent identifiers rule to include only those that can be used to recognize a user over time and across different unaffiliated websites. In this context, the FTC indicated, nodding to its 2012 Final Privacy Report, that the term “different” means either sites or services that are unrelated to each other, or where the affiliate relationship is not clear to the user. In the Privacy Report, this was explained as meaning corporate affiliates are third parties to one another unless their relationship is clear to consumers based on indicia such as branding. It excludes “hidden” affiliate relationships not clear to consumers, such as the FTC’s example of an online publisher that offers content to consumers through a website and an ad network that invisibly tracks consumers’ activities on the site. But where such connections are clear, the use of a persistent identified among a family of websites would not satisfy the “across different unaffiliated websites” requirement.
But at bottom, any information that identifies a particular personal computer or handheld device can be included within the definition of “personal information,” implicating the notice and consent requirements when linked to children under the age of 13. This marks a considerable expansion of the definition, and thus of the scope of the COPPA rules, and has the potential to significantly impede, if not ban, the use of behavioral targeting and other advanced advertising techniques by websites directed to children. In addition, even if the personal information that a website directed to children collects involves persistent identifiers used for internal operations, thereby avoiding any parental notice or consent requirements, other elements of the COPPA Rule apply, such as its requirements for data security, deletion, etc.
Definition of “Collects or Collection”
The FTC amended the definition of “collects or collection” that applies to children’s online personal information, and which thus partly controls whether COPPA obligations attach, so that in addition to applying when a website “requests” that children submit personal information online, the definition also applies if a site “prompts” or “encourages” such submissions. This intends to clarify, the FTC explained, its precept that websites that offer a field or open forum allowing children to enter personal information are not shielded from COPPA merely because the entry of personal information is not mandatory. Operators of websites covered by COPPA therefore must have a system to provide notice to parents and obtain consent from them to deal with the moment when personal information is “gathered” at a site. In addition, the FTC streamlined the “collect or collection” definition so that it more clearly covers all “passive tracking of a child online,” regardless of the technology used.
At the same time, the FTC relaxed somewhat its exception to the “collects or collection” definition, with respect to when a website (or online service) operator does not “collect” information within the meaning of COPPA if it deletes the individually identifiable information from both children’s postings before they become public, and from the site’s records. Under the former Rule, this exception applied only when the operator achieved a “100% deletion” standard. But the amended Rule requires only “reasonable measures” to delete all or virtually all of the information. In addition, the FTC reaffirmed that COPPA’s scope is limited to information collected online, and does not govern offline collection.
Definition of “Operator”
The COPPA definition for “operators” of websites or online services directed to children determines who must give notice and obtain parental consent when a child’s personal information is collected. An “operator” is anyone who operates a website or online service and collects or maintains personal information from or about users or visitors, or on whose behalf such information is collected or maintained. The Rule imposes “strict liability” on operators for COPPA compliance.
As to when personal information is collected “on behalf” of an entity, making it an operator, the FTC added a proviso that personal information is collected or maintained on behalf of an operator if it is collected or maintained by an agent or service provider of the operator, or the operator benefits by allowing such other person to collect personal information directly from users. The FTC offered, as examples of such benefits, instances where a content-providing site or service integrates plug-ins to enhance the functionality or content of their properties, to gain greater publicity through social media in an effort to drive more traffic, or to obtain direct compensation or increased revenue from advertising networks or other plug-ins. These benefits, the FTC explained, are not merely incidental, but may be crucial to a site’s continued viability.
The FTC also decided to maintain “strict liability” for child-directed content providers that allow other online services to collect personal information through their sites. In other words, operators of sites directed to children must presume any information collected from users includes children under age 13, though the FTC newly provided some flexibility based on age-gating, as discussed below. The same strict liability standard would apply to a general audience content provider that allows a plug-in to collect personal information from a specific user when the provider has actual knowledge the user is a child. In addition, when a behavioral advertising network offers age-based advertising segments that target children under 13, that portion of its service becomes an online service directed to children.
The Rule extends COPPA requirements to operators of ad networks or other downloadable plug-ins as “co-operators” only if they have actual knowledge that they are collecting personal information through a site directed to children. Although “actual knowledge” is a subjective, fact-based standard, it will be met when the third party receives direct communication from a child-directed site that it is collecting information through that site, or if the third party recognizes the child-directed nature of its content. However, the FTC did not rule out that other facts, taken together on a case-by-case basis, could suffice to establish actual knowledge.
The FTC later clarified, after the revisions were adopted, that determining when a third party has actual knowledge, based on a representative recognizing the child-directed nature of content, does not mean an email to any inbox at the company constitutes actual knowledge – rather, it has to be something that reached a person with sufficient responsibility for the online activity – though a sufficiently detailed complaint from an advocacy group might suffice. Nor is such actual knowledge chargeable based on just a website’s URL. In this regard, the third party does not have a duty to reach out to first parties to gain actual knowledge of the first party’s practices.
“Website or Online Service Directed to Children”
Another key trigger for COPPA compliance obligations, in addition to sites or services having actual knowledge they are collecting children’s personal information, is whether they are “directed to children.” This is based on a totality of circumstances that examines the site’s or service’s subject matter, content, age of models, verbiage, and whether ads promoting or appearing on the site or service are directed to children. Competent demographic data on users can also be considered. And now, under the Rule amendments, the “directed to children” criteria also include the type of musical content, and the presence of child celebrities and/or celebrities who appeal to children. However, the FTC has said post-adoption of the rule changes that it did not intend, with the revision, to expand the universe of COPPA-covered websites, but rather to simply capture and make more explicit criteria that always could have factored into “directed to children” analyses.
If a website is deemed to be directed to children, the designation generally means treating all visitors to such sites as children, and providing notice and/or getting consent for the collection of personal information as the Rule requires. However, the new amendments now allow a subset of sites that are “directed to children” the option of not treating all visitors as children, if the site does not target children as its primary audience and opts to use age-screening to apply COPPA’s safeguards to only those visitors who self-identify as younger than 13.
The net effect of these further revisions is that sites and services at the far end of the “child-directed” continuum—i.e., those that knowingly target, or have content likely to draw, children under 13 as their primary audience—must still treat all users as children, and provide notice and obtain consent before collecting personal information. Conversely, sites and services with child-oriented content that target mixed audiences, where children under 13 are likely to be an over-represented group, will not be “directed to children” if, prior to collecting any personal information, they age-screen all users.
In making this change, the FTC acknowledged children sometimes misrepresent their ages to access websites they wish to visit, but implicit in that acknowledgement is that there is no practical way to overcome this while still having a workable rule. At the same time, however, shortly after the rules were adopted, and before they appeared in the Federal Register, FTC personnel warned that under the new rules the agency would likely take a harder stance against mixed-audience sites that don’t adequately protect information collected from children.
The FTC requires that parental notice be “prominent and clearly labeled” and that it be posted on the website’s home page as well as at each location where personal information is collected from children. As for notice content, the rules require website operators to provide their contact information, including name, address, telephone number and email address. While the FTC considered requiring contact information for all operators collecting, using or disclosing personal information on a website or online service, it opted not to change the requirement. Thus, the primary operator of the site or service is the only entity required to provide contact information, along with a simple list of all other operators collecting, using or disclosing the personal information.
For each direct notice provided to parents, the FTC required that operators provide the following: the items of personal information already obtained from the child (generally, contact information only); the purpose of the notice; actions that the parent must or can take; and the operator’s use of the information collected. Each notice must also contain a hyperlink to the operator’s information practices. The Rule revisions streamline the notice content requirements to specify a “simple statement” of what personal information the operator collects from children (including whether the site or service enables a child to make personal information publicly available), how the operator uses the information, and the operator’s disclosure practices for it. Operators are no longer required to state that they will not condition a child’s participation on the provision of more personal information than is necessary, though that underlying substantive limitation remains in place.
Parental Consent Mechanisms
Critical to the operation of COPPA is the mechanism by which website and online operators receive “verifiable parental consent.” These have included providing a consent form to be signed by a parent and returned by postal mail or fax, requiring the parent to use a credit card in a transaction, having the parent call a toll-free number staffed by trained personnel, digital certificates that use public key technology, and using e-mail accompanied by a PIN or password. The Rule also provided for a method called “email plus,” usable when personal information is collected from children for internal purposes only, which entails using an email address plus one other piece of information, such as an address or telephone number.
The FTC added to these consent mechanisms so that the Rule now allows electronic scans of signed consent forms, videoconferencing, collection of parents’ government-issued IDs (provided the operator deletes the information immediately upon completing parental verification), and, where the parent’s involvement includes a monetary transaction, the use of a credit card, debit card, or other online payment system that provides notification of each discrete transaction to the account holder. The FTC considered but decided against eliminating “email plus” as a measure that has outlived its usefulness as a reliable means of consent, though it voiced its disdain for the method and concern that it continues in heavy use. The FTC thus issued an underscored reminder that email plus is available only for collecting personal information for internal use. At the same time, the FTC declined to allow parental consent features in game consoles to become a means of obtaining parental consent.
The FTC adopted a new exception to the need to obtain parental consent, in order to allow operators to obtain parental contact information from children solely for the purpose of voluntarily notifying the parent of a child’s participation in a service that does not otherwise collect, use or disclose children’s personal information. It also codifies into the exceptions to parental consent the above-described collection of a persistent identifier – and no other personal information – where used solely to support a site’s or service’s internal operations. Existing exceptions continue to allow communication with children to initiate the parental consent process, to respond to the child, and to protect the child’s safety or the website’s integrity.
Confidentiality and Security Requirements
COPPA requires that operators “establish and maintain reasonable procedures to protect the confidentiality, security, and integrity of personal information collected from children.” The new rules extend that provision to require operators who release children’s personal information to any third party to inquire into that entity’s data security capabilities, and to obtain by contract or otherwise assurances about how the third party will treat the personal information it receives. As a corollary, the FTC adopted a new rule requiring operators to retain children’s personal information for only as long as is reasonably necessary to fulfill the purpose for which the information was collected, and to delete personal information no longer needed, taking “reasonable measures to protect against unauthorized access to, or use of” the deleted information.
Safe Harbor and Pre-Approval for New Consent Methods and Activities in Support of Internal Operations
COPPA established a “safe harbor” program allowing participants in FTC-approved self-regulatory programs to be deemed “in compliance” with COPPA. TRUSTe and CARU (Children’s Advertising Review Unit) are examples of organizations that operate FTC-approved safe harbor programs. The FTC made several changes to strengthen the reliability and accountability of safe harbor programs. Specifically, the Rule requires (a) that safe harbor programs conduct an annual, comprehensive review of their members’ information practices; (b) that new safe harbor program applicants explain their business model and technological capabilities and mechanisms for assessing operators’ fitness for membership in the safe harbor program; and (c) that safe harbor programs conduct an independent audit of their programs every 18 months and submit the results to the FTC along with an aggregated summary of any disciplinary actions taken against member operators within that 18 month time period.
To the extent new rules require existing safe harbor programs to modify their guidelines, such modifications must be made within 60 days after publication of the final rule to avoid potential revocation of safe harbor status. Annual reviews are required starting July 1, 2014, one year after the effective date of the new rules. In addition, the FTC adopted new rules to establish streamlined methods of qualifying new means of obtaining verifiable parental consent, and as noted above, to include new activities in the definition of “internal operations” that do not require consent for collection and use of personal information.
* * * *
As noted at the top, the Rule amendments take effect starting July 1, 2013. This means that, over the coming months, operators of websites and other online services that collect (and use and/or disclose) children’s personal information, and those that collect personal information that are – or may be – directed toward children under 13, should be re-examining their practices in the wake of the new Rules.