Have you ever looked at a product online and realized it was following you around the internet? Have you ever visited a different website and seen the item you were just thinking about purchasing? These friendly reminders are due to cookies–small text files stored on your browser when you visit or interact with a website or advertisement. Continue Reading CCPA’s Uncertain Effect on Digital Advertising
In this podcast, Ballard Spahr consumer financial services partner Chris Willis talks with Scott Ferris, CEO of Attunely, a provider of machine learning (ML) and artificial intelligence (AI) technology to the debt collection industry. The podcast addresses how changes in consumer behavior have impacted collections, technology’s role in collections, how ML/AI can improve profitability, and impediments to adopting ML/AI. Phil Yannella, Leader of Ballard’s Privacy & Data Security Group, also discusses how the GDPR, CCPA and other US state privacy laws attempt to regulate ML/AI. Check it out!
Following on the heels of a few relatively small HIPAA settlements, the U.S. Department of Health and Human Services Office of Civil Rights (OCR) announced that it has imposed $2,154,000 in civil monetary penalties against Jackson Health System in Florida for its failure to meet HIPAA privacy and security requirements. The OCR announcement and accompanying information detail violations that included:
- The unauthorized access by an employee to the records of more than 24,000 patients over a five-year period (the employee admitted to selling protected health information of more than 2,000 patients for purposes of identity theft).
- The unauthorized access by staff members to protected health information about a professional athlete who received services at the health system (with some of the information revealed on public media).
- The loss of certain patient records.
- The failure to conduct adequate risk assessments, undertake appropriate measures to manage risks that were identified, and review logs that might have shown inappropriate access to information.
- The failure to implement and maintain adequate policies and procedures to respond to breaches and the failure to report breaches timely and fully.
Significantly, this case did not involve a settlement between OCR and the health system. The health system did engage with OCR in the course of the investigation, but ultimately chose to accept the civil monetary penalty. As a result, the materials do not include a specific corrective action plan for the health system to follow under OCR supervision. The materials do identify measures that the health system has undertaken to improve its privacy and security programs.
Settlement agreements typically provide limited information. By contrast, the notices published in this case provide not only details about the health system’s violations, but information about how OCR determined the amount to assess in civil monetary penalties. It considered various factors, including the nature and extent of the violations and the harm resulting from those violations, the history of the health system’s compliance, and the health system’s financial condition and cooperation in the investigation. OCR also took into account the health system’s mitigating and corrective actions.
Notwithstanding the size of the civil monetary penalty, it could have been larger. OCR chose to group violations into three broad categories, relating to failures in the security management process, information access management, and the provision of notice to HHS. It viewed the first two of these failures as attributable to reasonable cause. New limits cap penalties for any one type of violation arising from reasonable cause at $100,000 per year. As a result, most of the civil monetary penalty in this case is attributable to the health system’s failure to provide OCR with timely and accurate notice of a breach caused by a loss of paper records. OCR viewed this failure as one of willful neglect, for which penalties were capped at $1.5 million, even though this violation was seen as lasting only 31 days.
The materials published by the OCR serve as a warning about issues that might arise, particularly with regard to the implementation of policies designed to prevent and detect HIPAA violations. They also provide insight into how OCR is prepared to both impose significant civil monetary penalties and temper the amount of those penalties, even in situations that do not involve a formal settlement agreement.
For businesses, one of the more worrisome scenarios under the CCPA occurs when they mistakenly provide personal information of a consumer to the wrong party in response to a consumer request, whether because of fraud or simple mistake. Because the definition of data breach under the CCPA is very broad, the unauthorized sharing of personal information with the wrong party could theoretically give rise to a civil cause of action with statutory penalties of $100-$750, per consumer. As a result, businesses have been anxiously waiting to see how the proposed Regulations would address the consumer verification process.
The good news for businesses is that the proposed Regulations provide significant detail concerning the verification process, and those details will likely assuage the concerns of many businesses about potential litigation.
As a general matter, the proposed Regulations require that businesses verify consumers wherever possible using personal information collected from the consumer (or use a third-party identification service that complies with same requirements). Businesses should use reasonable security measures to detect fraudulent identity verification procedures and prevent the unauthorized access to or deletion of a consumer’s personal information.
Password Protected Accounts
For password-protected accounts, the proposed Regulations allow businesses to verify the consumer’s identity through its existing authentication practices if those practices are otherwise consistent with the CCPA regulations. Businesses must also require that consumers making requests through password-protected accounts re-authenticate themselves before responding to a deletion or right to know request. If the business believes that there is fraudulent or malicious activity on a password-protected account, it may require additional verification procedures to confirm the consumer request is authentic.
Two-Tier Verification For Non-Password Protected Accounts
The proposed Regulations outline a two-tier verification process for non-password protected accounts. This process requires that businesses verify requests to know categories of personal information to a “reasonable degree of certainty.” To meet this standard, businesses could match two (2) pieces of consumer provided personal information with personal information maintained by the business. For requests to know specific pieces of information, businesses must verify the consumer to a reasonably high degree of certainty, which can be accomplished by matching three (3) pieces of consumer provided personal information with personal information retained by the business. The proposed Regulations provide businesses with discretion for verifying requests to delete, depending on the sensitivity of the personal information. The table below lays this out.
|Abbreviated Right to Know (Categories of Personal Information)||Right to Know Specific Pieces of Information||Right to Delete|
Using information provided by consumer, verification must be to reasonable degree of certainty, which may include matching at least two data points provided by the consumer.
|All deletion requests require verification at the time of the request, and re-verification before any data is deleted.
Business may use its discretion, based on sensitivity of data, whether to use two or three-step verification.
E.g., deletion of browsing history may only require reasonable degree of certainty whereas deletion of family photos may require reasonably high degree of certainty.
Using information provided by consumer, verification must be to reasonably high degree of certainty, which may include matching at least three data points provided by the consumer and obtaining a signed declaration under penalty of perjury that the requestor is the consumer whose personal information is the subject of the requests.
The proposed verification process, especially for a right to know specific pieces of information collected about the consumer, sets a high standard—particularly insofar as the proposed Regulations discourage businesses from using sensitive information such as SSN for matching purposes (as discussed below). The declaration envisioned by the proposed Regulations adds an additional hurdle for consumers.
Businesses will have to make a determination of what pieces of personal information to use for matching purposes based on what information they are holding for consumers, and the sensitivity of such information. One likely consequence, however, is that the added verification procedures will reduce the number of verifiable requests to know specific pieces of information as well as requests to delete that businesses must honor. Don’t be surprised if privacy advocates decry the heightened verification procedures in the upcoming Public Comment period.
Shielding Sensitive Data From the Verification and Response Process
The proposed Regulations directly address the concern discussed above about potential civil causes of action stemming from mistakes made during the consumer request process by explicitly prohibiting businesses from providing in response to a request to know the following sensitive information:
- social security number
- driver’s license number
- state identification number
- medical and health information
- financial account number
- account passwords, or
- security questions and answers.
The proposed Regulations, furthermore, prohibit businesses from requesting such information in the verification process “unless necessary.”
The CCPA’s private right of action only applies to unauthorized access and theft, exfiltration or disclosure of personal information as defined under the California breach notification statute—which is notably narrower than the definition of personal information under the CCPA and similar to the sensitive data listed above. The effect of the proposed Regulations, then, is largely to prohibit businesses from unnecessarily collecting or disclosing the kinds of data that might trigger a cause of action under the CCPA in the event of a breach.
Use of Authorized Agents to Make Consumer Requests
The CCPA allows consumers to use authorized agents to make consumer requests, increasing the potential for fraud where tricksters impersonate authorized agents or even use bots to scam businesses into providing consumer information. The proposed Regulations allow businesses to guard against this by requiring that such agents produce a signed authorization from the consumer. Businesses can also require that consumers using agents separately verify their identify to the business to prevent fraud
Requirements Where a Business Cannot Verify a Consumer
The use of a categorical disclosure may be particularly useful for a business that receives requests to know from website visitors who do not have accounts open with the business and otherwise haven’t provided any other information that the business can use to verify their identity.
To the surprise of some, the proposed CCPA Regulations issued last Thursday don’t address many of the well-discussed ambiguities under the law (such as what “valuable consideration” means in the context of a sale of personal information). Rather, the proposed Regulations address a number of technical, nut-and-bolt type compliance issues concerning how businesses must make required privacy disclosures, provide opt-outs notices, verify and respond to consumer requests.
We’ll discuss the issues the proposed Regulations don’t address in a future post. But for now, we turn our attention to specific sections of the proposed Regulations, highlighting any new legal compliance issues the proposed Regulations raise.
We’ll start with Article 2, which covers “Notices to Consumers”. The CCPA requires that businesses disclose to consumers at or before the time of collection the categories of personal information to be collected from consumers and the purposes for which the personal information will be used. Businesses are prohibited from collecting or using personal information that is not disclosed prior to collection, and must obtain explicit consent from the consumers for any new use that was not previously disclosed.
Online v. Offline Disclosures
The first point to note about the Notice provisions of the proposed Regulations is that they explicitly differentiate between online and offline privacy notices. This clarifies an ambiguity in the CCPA itself, which requires notice at or before collection but doesn’t explain how businesses can provide notices online and offline.
Here’s what the proposed Regulations say about the mechanics of providing Notice:
Notice of information collected from consumers and business or commercial use of such personal information
|Opt Out||If business sells personal information, it must furnish a “Do Not Sell My Personal Information” or “Do Not Sell My Info” link in the Notice.||Business must provide URL for webpage to which the “Do Not Sell My Personal Information” or “Do Not Sell My Info” links direct users.|
It is likely that many businesses had not considered online accessibility issues before the proposed Regulations came out. Many businesses have gone through an ADA website accessibility analysis in the past. Those who have not should consider doing so now.
New Attestation Requirements
The fourth and final provision of Article 2 that’s noteworthy appears in section (d), which makes clear that a business that does not collect information directly from consumers does not need to provide a Notice of collection. Before selling any such personal information, however, the business must either:
- Contact the consumer directly to provide the required Notice, and Notice of Opt-Out; or
- Contact the source of the personal information and confirm the source provided the required Notice and obtain a signed attestation describing how the source gave the Notice at collection along with an example of the Notice.
The upshot of this new requirement is that businesses that either sell or share data with third party data brokers or lead generators, or others in the business of selling data, can assume that they will see requests from these third-parties for attestations. For such businesses, it may make sense to consider the content and format of the attestations as well as how they might automate the process of providing the attestations, if the volume is likely to be significant.
In our next post, we will address the new verification procedures outlined in the proposed Regulations.
The California Attorney General’s Office released its long-awaited proposed CCPA Regulations this afternoon. The proposed Regulations are 24 pages long, and address a number of important technical compliance issues including how businesses should:
- provide just in time notice to consumers of personal information collected;
- provide notice to consumers of the right to opt out of the sale of personal information;
- provide notice to consumers of financial incentives;
- provide methods for consumers to submit requests to know and requests to delete their personal information;
- respond to consumer requests to know and requests to delete their personal information
- respond to consumer requests to access or delete household information;
- respond to requests to opt-out;
- respond to requests to opt-in after consumers exercise their right to opt out of the sale of personal information; and
- verify consumer requests.
The AG’s office also released a 97 page Initial Statement of Reasons (including appendices).
Ballard’s Privacy & Data Security lawyers are carefully reviewing the proposed Regulations. We will post our thoughts on the effect of the proposed Regulations, what they mean from a compliance standpoint, what issues the proposed Regulations fail to address, and what’s next for the CCPA in the coming days.
The perplexing question of what U.S. companies must do to comply with EU “cookie” law became slightly more clear with the recent decision of the European Court of Justice (CJEU) in Planet49 GmbH, but numerous questions still remain. A main source of confusion about cookies is the interplay between two EU privacy laws, the ePrivacy Directive and the GDPR. The former governs, among other things, the placement of cookies and marketing pixels on the browsers of website visitors and the latter governs the subsequent processing of personal data, which in many cases includes cookies. Some cookies, in other words, are subject to the ePrivacy Directive but not the GDPR. Another complication is that the ePrivacy Directive does not have an extra-territorial effect whereas the GDPR does have such an effect.
Many privacy professionals had hoped that the CJEU’s ruling in Planet49 would provide some much-needed clarity to a muddled legal picture. And it does, sort of.
The case involved participation in a lottery organized by Planet49 GmbH, an online gaming company. To enter the lottery, internet users were prompted to enter their personal information, then presented with two checkboxes. The first required the user to agree to be contacted by other businesses for promotional offers. The second checkbox, which contained a pre-ticked box, required the user to consent to cookies. In order to participate in the lottery, the first checkbox needed to be ticked.
The first question referred to the CJEU concerned whether the use of a pre-ticked box was sufficient to obtain valid consent for placing cookies on a user’s device. The second question referred to the CJEU was whether service providers need to give users information specifically about the duration of the operation of the cookies and access by third parties.
How Planet49 Establishes Some Clear Guidelines
The CJEU’s recent ruling in Planet49 helps to clarify some of the rules governing the placement of cookies. First, the CJEU ruled that Planet49’s use of pre-ticked boxes is not a sufficient basis to establish consent. The writing had been on the wall for the use of pre-ticked boxes even before the GDPR became effective like other references, so this part of the ruling is not surprising.
What makes the ruling significant is the Court’s finding that consent requires some action on the part of the user. Although the ruling technically addresses consent under the ePrivacy Directive, the CJEU’s ruling suggests that inferring consent from passive activities, such as the continued browsing of a website, might not meet the GDPR’s more exacting “affirmative consent” standard. This finding aligns with recent guidance from the ICO, CNIL and the German Data Protection Authority (DPA), who have all issued guidance aligning consent under the ePrivacy Directive with the GDPR standard and have explicitly stated that the continued browsing of a website alone does not constitute consent. The CJEU did not go quite this far, but the days may be numbered for cookie banners that infer consent from continued browsing.
Second, the ruling makes clear that data controllers must gather consent for the placement of all non-essential cookies on a user’s device. This includes analytic cookies, which are commonly used by most companies with a website.
Lastly, the CJEU ruling requires that data controllers disclose the duration for cookie retention as well as the sharing of cookies with third parties in order to satisfy the ePrivacy Directive’s requirement that consent be “freely given specific and informed indication of the user’s wishes.” The ruling does not state what the maximum retention period for a cookie should be, but some EU regulators have suggested retention periods in recent guidance.
What the CJEU Ruling Doesn’t Resolve
For U.S. companies with physical operations in the EU, the CJEU ruling does not address a number of thorny issues. In particular, one open question is what, if any, user actions short of physically clicking an “Accept” button might constitute a valid cookie consent. While continued browsing might not be sufficient active to establish consent, would clicking out of a cookie banner?
The subsequent processing of tracking cookies for the placement of targeted ads is another issue that remains muddled in the wake of the CJEU opinion. The ICO has taken the position that a data controller cannot rely on legitimate interests as a basis for subsequent processing of cookies, particularly tracking cookies. The CNIL and the German DPA have not gone so far as the ICO and appear to leave open the possibility that legitimate interests may be permissible for subsequent processing of cookies.
For U.S. companies that don’t have an establishment in the EU, compliance is even more complicated insofar as these companies may be subject to the GDPR but not the ePrivacy Directive. The GDPR only governs the processing of cookies or other online identifiers that gather or contain personal information whereas the ePrivacy Directive covers the placement of any cookie or file on a user’s browser. Thus, for U.S. companies that don’t maintain an EU establishment, it remains unclear whether the guidance of EU data regulators regarding analytic cookies, for example, applies.
The bottom line is that for U.S. companies doing business in Europe, the CJEU’s recent ruling provides some important guardrails useful for fashioning cookie banners and policies, but numerous questions remain unresolved. Until an ePrivacy Regulation is released, U.S. companies will likely follow the proverbial herd, trying their best to hide in a crowd of other companies also struggling to understand where the lines lay.
Remarks Focus on Account Takeovers, BEC Schemes, Beneficial Ownership, Technological Innovation and SARs
FinCEN Director Kenneth A. Blanco delivered prepared remarks on September 24 at the 2019 Federal Identity (FedID) Forum and Exposition in Tampa, Florida.
Director Blanco summarized the topics of his remarks by stating the following:
- First, I would like to tell you a little about FinCEN. Who we are, what we do, and why I am here to speak with you today.
- Second, I will speak to how illicit actors are leveraging identity. Specifically, I will highlight some of the trends FinCEN is seeing in how criminals exploit and compromise identities.
- Third, I will discuss how we use identity to protect our national security and keep our communities and families safe from harm.
As to the “who” and “what” of FinCEN, Director Blanco emphasized the agency’s role as “Administrator of the Bank Secrecy Act” and “THE Financial Intelligence Unit” of the U.S. Director Blanco went on to describe current developments in how “identity” – described by Director Blanco as “who we are legally” – is employed in the financial sector and government. Such developments are “critically important,” in part because “the features that make identity information valuable to companies also make these data stores high value targets for criminals and other bad actors, including terrorists and rogue states.”
Director Blanco next addressed the abuse of personally identifiable information by means of “account takeover,” which involves the targeting of customer accounts to gain unauthorized access to funds. He noted that FinCEN receives approximately 5,000 account takeover reports each month (totaling about $350 million), but that this figure amount merely reflects “attempts” and not actual losses. Director Blanco further noted that, “Criminals often acquire these leaked credentials through hacks, social engineering, or by purchasing them on darknet fora to facilitate the account takeover. Depository institutions, such as banks, are the most common targets given their high numbers of customer accounts, but institutions like insurance companies, money services businesses, and casinos, and of course their customers, are also affected.” Director Blanco then called for improving “cyber hygiene” by, among other things, implementing strong authentication solutions (such as multi-factor authorization and authentication procedures for processing payments or allowing access to sensitive information). He reminded the audience that FinCEN held in July 2019 a FinCEN Exchange on business email compromise (BEC) fraud schemes targeting U.S. financial institutions and their customers, and that FinCEN had issued a July 16, 2019 Advisory on BEC fraud.
Separately, Director Blanco warned of bad actors who exploit weaknesses posed by the ubiquity of Social Security numbers (“SSN”). A FinCEN analysis of Suspicious Activity Reports (“SARs”) filed since January 2003 found more than 600,000 SSNs affiliated with identity theft reported from financial institutions, many of which were associated with more than one name. “That is mind-boggling, and it points to something wrong with how identity is being verified and authenticated across much of the financial system.”
Director Blanco then discussed the use of identity as a means to counter illicit activity. In doing so, he emphasized that beneficial ownership information is a critical issue whose “importance to our national security cannot be understated.” Notably, Director Blanco criticized the lack of an ability to collect identity information as a “dangerous and widening gap in our national security apparatus.” Although he praised the agency’s promulgation of the customer due diligence rule (a topic on which we have written extensively, see, e.g., here, here and here), he called for a separate rule to collect beneficial ownership information at the corporate formation stage. “To be sure, it is not that shell companies should not exist—it is just that the authorities should be able to know who owns and controls them when there is a legitimate law enforcement need, subject to appropriate information access safeguards. But currently, there is no federal standard requiring those who establish shell companies in the [U.S.] to provide basic, but critical information at company formation.”
Finally, Direct Blanco stated that FinCEN has strongly supported “responsible innovation” in the financial sector in regards to using technological advances to comply with BSA regulations. “Innovative indicators that reveal customers’ digital footprints and activities are extremely helpful to financial institutions in the conduct of their day-to-day business, including helping them understand customer activity and monitoring for suspicious activity.” He observed that FinCEN changed the SAR form in 2018 in order to allow for the reporting of up to 99 technical indicators, such as IP addresses, MD5 hashes, PGP keys, and device identifiers.
Over the past several years, state legislatures have become more aggressive in passing laws to protect consumers’ digital rights. The promulgation of state data security and privacy laws, such as the California Consumer Privacy Act, is a prime example of this trend. Another less publicized example of state oversight of online activities is legislation regulating automatic renewals, which have become very common and present new and little appreciated regulatory and litigation risks. Here’s a quick primer on these laws.
As background, automatic renewals refer to the business practice of subscribing a customer to receive a product or service and billing customers periodically for products and services without needing to obtain their express consent before each charge. Automatic renewals can benefit both customers and businesses; customers enjoy having their favorite products or services delivered to them automatically and businesses benefit from steady delivery of their products and services. On the other hand, regulators remain concerned that these automatic renewals can be misused by online retailers, publications and service providers, who may not always provide consumers with adequate disclosures or provide an easy mechanism to cancel their subscriptions before being charged again.
On the federal level, internet-based automatic renewals are regulated by the Federal Trade Commission (FTC) under the Restore Online Shoppers’ Confidence Act (ROSCA). This law requires clear disclosures of material terms, informed consent before obtaining financial information to process a purchase, and a simple mechanism to cancel the charges. Violations of ROSCA are categorized as unfair or deceptive acts or practices under the Federal Trade Commission Act. The FTC has become more aggressive in policing ROSCA, recently settling charges against Hardwire Interactive Inc. for $3,000,000.
On the state level, 26 states have implemented some form of automatic renewal laws. States that recently adopted these laws include California in 2018, and Vermont, Virginia, Washington, D.C. and North Dakota, in 2019. The North Dakota law is the most recent and provides a private cause of action for consumers injured by illegal renewals. Virginia also provides a private right of action for injured consumers.
In California, the law requires disclosures to customers for automatic renewals, free gifts, and trials as well as to offer a way for customers to cancel their subscriptions online. Vermont’s automatic renewal law requires that in addition to accepting the contract, the customer must also take affirmative action to opt into automatic renewal provisions of the contract. In Virginia, businesses must obtain a customer’s affirmative consent to the automatic renewal terms prior to charging the customer. In D.C., customers need to opt into the subscription after the end of their free trial, requiring the business to go back to the customer and ask for their consent.
These laws require businesses to restructure the way their current automatic renewal processes work. Businesses must take into account the additional consent requirements and update their systems so that customers are not automatically billed until the business receives all the opt ins required to process the transaction. Also, because most businesses conduct business in more than one state, the automatic renewal process needs to be compliant for every state in which the business operates.
The state laws also provide regulatory penalties for non-compliance. In October 2018, Spark Networks, the parent company of dating websites JDate and Christian Mingle, settled an enforcement claim brought by the California Attorney General for $1,500,000 for automatic renewal violations. In addition to fines, some states give customers additional rights regarding automatic renewals conducted in violation of the law. In California, if the business sends a product without following the requirements of the law, the customer may keep the product as an unconditional gift.
We are in the midst of a period of increased regulatory scrutiny of businesses’ online activities. If your business uses automatic renewals to deliver products and services to customers, you should consider reviewing your practices to ensure compliance with these new laws.
On September 13, 2019—the last day of the legislative session—California lawmakers approved five amendments intended to clarify the scope of the California Consumer Privacy Act (the “CCPA”), but rejected several industry-backed proposals that would have exempted personal information used for targeted advertising and loyalty programs.
Five amendments passed: AB 25, 874, 1146, 1355, and 1564. As we have noted in prior posts, the version of AB 25 that ultimately made it through the legislature had changed from a more business-friendly exclusion for certain employment-related information to a compromise bill, whereby employers must still inform employees of the types of information they are collecting and the reason for doing so. AB 25 also subjects employers to the private right of action with statutory damages in the event of a data breach, albeit only as to “personal information” as defined in California’s data breach notification law. AB 25 has a one-year sunset provision, after which employee personal information will be treated the same as consumer personal information without further legislative action or regulatory guidance.
AB 1355 is also particularly important as it excludes from consumer personal information: (1) consumer information that is deidentified or aggregated; and (2) personal information gathered in the context of a business-to-business transaction. While the latter exclusion has a one-year sunset provision, this exclusion provides a significant boon to businesses that engage primarily in B2B transactions, but were nonetheless previously concerned that they may hold significant amounts of personal information under the CCPA’s broad definition.
Notably, the legislature did not pass AB 846, which would have allowed companies to collect personal information to offer loyalty programs without worrying that the practice was discriminatory under the law. The legislature also rejected proposals backed by the California Chamber of Commerce and the Internet Association—which includes Google, Facebook, and Amazon in its members—which would have increased exclusions relating to targeted advertising and fraud detection, as well as expanded the definition of “deidentified.”
Although this year’s legislative session is complete, there is still a chance that the Attorney General’s forthcoming regulations could alter the CCPA’s scope and application. The proposed regulations are expected to be issued later this month or in October, followed by a comment period prior to finalization.
However, assuming the Governor signs passed amendments, companies now know the version of the CCPA which will go into effect on January 1, 2020, and should prepare accordingly.