With one eye on the New York Department of Financial Services (“NYDFS”) cybersecurity rules and the other on two of its own Commissioners who dissented, the Federal Trade Commission (“FTC”) has proposed a sweeping overhaul to its Gramm–Leach–Bliley Act (“GLBA”) data security standards for financial institutions subject to its authority (“Safeguards Rule”).[1] Specifically, on March 5, 2018, the FTC released a draft notice of proposed rulemaking (the “Proposed Rule”) to amend substantially the Safeguards Rule. The Safeguards Rules has long been a hallmark of a reasonable approach to regulating data security, utilizing a risk–based framework and approach that is technology–neutral and that does not create a one–size–fits–all standard. This historical approach is premised on a financial institution maintaining an information security program, performing periodic risk assessments and including within the program administrative, physical and technical safeguards designed to protect against the identified risks. With the Proposed Rule, however, the FTC appears to be heading to the opposite end of the spectrum and would impose detailed and prescriptive security requirements, including broad obligations regarding the use of encryption and multi–factor authentication.
The FTC commentary accompanying the Proposed Rule asserts that the proposed changes are “informed by the FTC’s almost 20 years of enforcement experience.” Nonetheless, many of the specific requirements that would be imposed by the Proposed Rule track, almost verbatim, the controversial cybersecurity regulations promulgated by the NYDFS at the end of 2016 and that have just recently taken full effect (“NYDFS Cyber Rules”).[2] Given the depth and breadth of the FTC’s experience with information security issues, both under the GLBA and the FTC Act, as well as its role as the country’s most prominent and active enforcer regarding data security matters, it is striking to see the FTC rely so heavily on a rule issued by a state agency, particularly one that was significantly criticized for not being well thought out and for including unworkable and inflexible standards.
While the FTC states that it “does not believe that the proposed new requirements would require an overhaul of existing compliance programs,” if adopted in its current form, the Proposed Rule would in fact impose substantial compliance costs and challenges for covered financial institutions.[3] Without any meaningful indication that the Safeguards Rule is “broken” or not working as intended, the FTC’s proposal appears inconsistent with the current administration’s emphasis on regulatory relief and deregulation. Regardless, the FTC is likely to receive significant critique on the Proposed Rule, including the fact that the Proposed Rule would disincentivize covered financial institutions from approaching the complex issue of security in a thoughtful, risk–based and risk–targeted manner.
The following provides an overview of key aspects of the Proposed Rule. Comments on the Proposed Rule will be due 60 days after the Proposed Rule is published in the Federal Register.
Proposed Security Requirements
The Proposed Rule contemplates a number of changes to the Safeguards Rule. The most significant changes relate to the new prescriptive, security controls that the FTC would mandate that a covered financial institution include in its information security program.
The Proposed Rule would preserve the foundational obligation under the Safeguards Rule for a financial institution to develop, implement and maintain an information security program that includes administrative, technical and physical safeguards to protect “customer information.” Nonetheless, in a dramatic departure from the current risk–based approach, the Proposed Rule would mandate that a financial institution include a number of specific technical and administrative controls within its program, including controls required by the NYDFS Cyber Rules. Furthermore, while the proposed amendments still include the proviso that the information security program be based on a risk assessment and that safeguards be designed and implemented to control identified risks, the drafting is far from clear as to whether the specific enumerated controls must be deployed based on the risk assessment, similar to the NYDFS approach, or whether the controls are required irrespective of the risk assessment.
Because of the breadth of many of the specific technical requirements, the issue of whether the specific controls must be adopted based on a risk assessment (as opposed to mandated generally) will be a critical issue. In particular, borrowing from the NYDFS Cyber Rules, the Proposed Rule would require that an information security program address, among many others, the following controls:
-
Encryption. The Proposed Rule would require that a financial institution’s information security program “[p]rotect by encryption all customer information held or transmitted by [the financial institution] both in transit over external networks and at rest.” The drafting of this provision is not clear. For example, it is not clear whether the Proposed Rule would require encryption of all customer information that is stored or transmitted, or only when customer information is transmitted over external networks or stored at rest specifically. This encryption requirement also appears inconsistent with the FTC’s de facto standard established through guidance and enforcement actions, which is to securely store and transmit sensitive personal information in particular (as opposed to personal information generally).[4]
-
Multi–factor authentication. The Proposed Rule also would require that a financial institution implement “multi–factor authentication for any individual accessing customer information.” Standing alone, this requirement appears to be unworkably broad, if not impossible to meet. For instance, the definition of “customer information” includes records in any form, including paper. It is far from clear how the FTC envisions a financial institution deploying multi–factor authentication to control access to customer information in paper or even verbal form. Regardless, the Proposed Rule would track the NYDFS Cyber Rules by establishing that “[m]ulti–factor authentication shall be utilized for any individual accessing [the financial institution’s] internal networks that contain customer information.” Much like the proposed language relating to encryption, this drafting is not clear. More specifically, it is not clear whether the statement surrounding access to a financial institution’s internal network is intended to clarify and define the scope of the required use of multi–factor authentication, or if it is in addition to, or an example of, the requirement to use “multi–factor authentication for any individual accessing customer information.” The Supplementary Information accompanying the Proposed Rule, which appears to address multi–factor authentication only in the context of a consumer accessing his or her own customer information, does not offer any additional clarity.
-
Access controls on information systems. Separate and apart from multi–factor authentication, the Proposed Rule would require that a financial institution include access controls on its information systems in order “to authenticate and permit access only to authorized individuals,” and to periodically review these access controls.
-
Access restrictions. Continuing with the proposal’s significant focus on access, a financial institution would be required to “restrict access at physical locations containing customer information only to authorized individuals.”
-
Management. The Proposed Rule would require that a financial institution “[i]dentify and manage the data, personnel, devices, systems, and facilities that enable [it] to achieve business purposes in accordance with their relative importance to business objectives and [the financial institution’s] risk strategy.” This obligation is far from clear. In particular, the Proposed Rule does not state the purpose for which the litany of people and things (e.g., facilities) must be identified and managed. The Proposed Rule also does not make clear what type of “manage[ment]” would actually be required.
-
Audit trails. The Proposed Rule would require that a financial institution put in place “audit trails within the information security program designed to detect and respond to security events.” Much like with the NYDFS Cyber Rules, this obligation would confuse the process surrounding detecting and responding to security incidents, which is an active function, with an audit function, which is a retrospective function.
-
Other technical controls. The Proposed Rule would include a litany of other specific technical controls, including requirements surrounding application development, disposal, change management and monitoring.
In addition to the various technical controls that would be required by the Proposed Rule, the Proposed Rule also includes a number of specific administrative and assessment obligations, including training of personnel, oversight of vendors (consistent with the Safeguards Rule) and annual penetration tests and biannual vulnerability assessments (from the NYDFS Cyber Rules). The Proposed Rule also would require covered financial institutions to establish a written incident response plan addressing, among other things, internal processes for responding to security events and documentation and reporting regarding security events. Of note, however, is that the FTC did not propose to include an express breach notification obligation for incidents involving customer information, notwithstanding the fact that the federal banking agencies have had in place such obligations for more than 15 years under their own GLBA security rules. In the context of such a dramatic overhaul of the Safeguards Rule, this seems like an odd, even if potentially welcome, omission.
Similarly, unlike the NYDFS Cyber Rules, the Proposed Rule does not include an obligation to certify compliance with the Rule periodically to the FTC. Instead the FTC takes a different page from the NYDFS Cyber Rules and would require that a financial institution’s chief information security officer report at least annually, in writing, to the board of directors or equivalent governing body of the financial institution. The Proposed Rule states that the report must include, among other things, “[m]aterial matters related to the information security program, addressing issues such as . . . security events or violations and management’s response thereto.” In the Supplementary Information accompanying the Proposed Rule, however, the FTC states that the annual report “must identify all security events that took place that year.” A “security event” is defined broadly to include “an event resulting in unauthorized access to, or disruption or misuse of, an information system or information stored on such information system.” Given that the FTC has indicated that “all” security events must be reported to the board and the fact that a “security event” includes any unauthorized access to information systems, as well as the fact that the definition of information systems is not limited to systems that handle customer information (let alone sensitive customer information), these provisions could be read together to require extensive reporting to governance committees of immaterial matters that go well beyond what they would be able to meaningfully digest.
Dissenting Statement
As noted above, two FTC Commissioners, Commissioner Phillips and Commissioner Wilson, issued a dissenting statement in connection with the FTC releasing the Proposed Rule. This dissent deserves attention, particularly because it cuts to the heart of many of the issues associated with the FTC’s proposal, several of which we have already highlighted.
In particular, the dissenting Commissioners cited four concerns with the Proposed Rule:
-
It is overly prescriptive. The dissenting Commissioners expressed the belief that mandating prescriptive standards for all market participants is ill–advised and that “[t]o the extent that the [FTC] thinks it is appropriate to elucidate the regulation’s reasonable care requirements, we have tools at our disposal — including speeches, testimony, analyses to aid public comment, information about the factors the [FTC] considered when closing investigations, and reports.” This comment highlights, among other things, that the FTC has many vehicles through which it can educate the marketplace, including covered financial institutions, regarding its expectations for security. While unstated, the FTC has “jumped” from high–level guidance to detailed and prescriptive requirements.
-
It is premature. The dissenting Commissioners note that the Proposed Rule is based in large part on the NYDFS Cyber Rules. In this regard, the dissenting Commissioners expressed concern that, given how recently the Rules were issued, there is not yet sufficient information about the “impact and efficacy of those [Rules], so whether to adopt a version of them at the federal level and whether that version should be a floor for or should preempt state–level rules seem like questions worthy of more study.” The dissent also highlights the fact that the issues of privacy and data security legislation are being actively debated in Congress, which is where that debate belongs.
-
There are costs associated with this new, inflexible approach. According to the dissent, the Proposed Rule would move away from the historic “flexible approach, appropriate to a company’s size and complexity.” It would create direct costs for such enhanced requirements, but the FTC’s record in issuing the Proposed Rule “does not demonstrate that those costs will significantly reduce data security risks or significantly increase consumer benefits.” While an astute point, this aspect of the dissent also raises the question of whether the FTC has identified a clear gap in security at covered financial institutions that needs to be addressed through a formal rulemaking of this magnitude.
-
It raises governance concerns. The dissenting Commissioners also expressed concern that, through the Proposed Rule, the FTC would be substituting its own judgment for private industry governance decisions, such as the appropriate level of board engagement, hiring and training requirements and accountability structures.
Next Steps
Given the breadth of the Proposed Rule as drafted, as well as the potential compliance and implementation challenges it would raise if adopted in its current form, financial institutions subject to the authority of the FTC with respect to GLBA data security will need to consider whether to make their voices heard during the comment period and, if so, to ensure that they convey the potential implications of this rulemaking for their businesses. It may be that, much like the evolution of the NYDFS Cyber Rule, which went through some fits and starts, the FTC may refine the concepts included in the Proposed Rule further before issuing a final rule. Regardless, this is an issue that financial institutions should closely monitor to see how it develops. In fact, given the FTC’s more general authority under the FTC Act, non–financial institutions should also monitor this rulemaking for signals related to whether this is the FTC’s new “norm” or expectations for the marketplace generally.
[1] 16 C.F.R. pt. 314.
[2] 23 N.Y.Comp. Codes. R & Regs. pt. 500. For more information on the NYDFS Cyber Rules, please refer to our various alerts regarding the Rules, including “One of a Kind: NYDFS Cyber Rule Finalized, Effective March 1, 2017” and “New York Cybersecurity Regulations: What Do They Mean and When Do They Mean it By?”
[3] The specific requirements of the Proposed Rule would not take effect until six months after the final rule is issued.
[4] See Start with Security: A Guide for Business (Lessons Learned from FTC Cases), available at: https://www.ftc.gov/system/files/documents/plain-language/pdf0205-startwithsecurity.pdf.
