Companies often claim that data is one of their most precious assets, but they rarely treat it as such. Mayer Brown Partner Dan Masur discusses The Big Data Paradox , and in this article, outlines a number of issues inherent in ensuring regulatory compliance when obtaining data from an outside source.
Companies obtain data from an increasing number of sources. Some of these sources are under contracts titled “data license agreements,” but most are under other types of agreements. Those other agreements might include subscription agreements, website terms of use, outsourcing agreements, purchase and sale agreements, alliance agreements and other commercial agreements.
Data acquired from third parties generally come with license and use restrictions and may come with restrictions that attach to personal data. In some cases, the license terms associated with the data are subject to significant negotiation. In other cases, however, a company accepts license terms with little thought as to whether they are aligned with the anticipated handling and use of the data.
To ensure compliance with applicable license terms, each item of licensed data must be linked to its source and to the specific terms on which the data was obtained. Unfortunately, data is often not tracked at all, or the data provenance is lost when the data flows into a database or from one database into another. The danger, of course, is that data is used in ways and for purposes not contemplated by the license. This can result in license breaches, privacy law violations, intellectual property violations and regulatory compliance failures.
Even keeping track of data can be challenging. Software often has a “software fingerprint” and may even be reporting on its use. By comparison, it may be costly or even impossible to identify all of the locations where licensed data is being stored or used. Thus, without advance planning and technology, it can be difficult or even impossible to demonstrate that a company’s data use is consistent with the terms of the applicable license grant and may expose it to significant liability in the event of an audit.
Tracking data provenance and its related restrictions is new to many companies, and, like many new areas, it requires that a company develop policies and procedures. When a company is licensing data from a third party, there are important considerations which, when properly managed, can lead to better data licenses. The following are important issues to be addressed when obtaining data from a third party.
The core provisions of a data license agreement define the data that is being licensed, including the manner and frequency with which the data will be provided/updated, how current the data will be (that is, whether the data will be provided on a “real-time” or close to “real-time” basis), the format in which the data will be delivered and the mechanism of delivery. Such terms may include the use of encryption and a secure delivery mechanism, designated communications technology platforms and specific hardware or software configuration requirements. These provisions vary from a general license that may be accessible to the licensee during the license term to a specific license — for example, to market data on specific assets within a specific time after the market event occurs.
The data license must also establish who is permitted to use the licensed data. For example, the license agreement may identify the people who are permitted to use the data or the devices on which the data may be used or may specify the maximum number of such users or devices. The licensee should be sure that any such restrictions are consistent with its anticipated use of the data. In addition, given the complex structures of many corporations, consider making clear that data use is not restricted to the entity executing the license and that the licensed data may be used by affiliates of that entity. Also, to the extent a company uses third-party contractors, it may be important to provide that the licensed data may be used by such third-party contractors in performing services on behalf of the licensee.
Finally, depending on the business model of the licensee, it may be important to provide that the licensed data may be accessed and used by regulators or customers of the licensee and its affiliates. Of course, it is also important to flow down to the affiliates, third-party contractors (and their subcontractors) and customers any license restrictions on the use of such data.
To the extent relevant, the data license agreement should also address the issue of exclusivity. Most data license agreements are nonexclusive, where the licensor has the same rights to the data as the licensee and can also license the data to other third parties. Less often, a licensee may require an exclusive license to the data, which will only grant rights to the data to the licensee, not allowing use or access by any other parties, including the licensor. A sole license is another option. A licensee may seek a sole license if it does not want the data to be licensed to other third parties, but to allow the licensor to continue to access and use the data.
In some cases, data is licensed for a specific purpose and only for that purpose. For example, in the case of a bank, a customer may provide data for the purpose of opening and maintaining an account, obtaining a mortgage or other loan, engaging in a corporate transaction, facilitating the completion of required “know-your-customer” checks, etc. However, in many cases, the data finds its way into other databases, where it is unwittingly used for new or different purposes. It is thus important for the licensee to seek to include in the data license (which, in this example, might be a customer agreement) all of the possible purposes for which the data may be used, including, to the extent possible, possible future uses. If the purpose clause is not as general with regard to those possible future uses, compliance processes are needed to avoid a possible license breach.
For companies that operate in many locations, it is important to focus on where the data can be stored, accessed and used. For example, the proffered data license may limit storage, access and use to the United States. If storage, access or use of the data outside the United States is contemplated now or may be in the future, make that clear in the license agreement.
Given the proliferation of data protection laws and the current focus on data privacy and cybersecurity, it is important to address in the data license the nature and sensitivity of the data to be provided, the steps the licensee is obligated to take to protect the data and the licensee’s potential liability if a data breach occurs.
Licensors often seek to disclaim any representation or warranty with respect to the completeness, accuracy, timeliness or utility of the licensed data. A licensee may see the following disclaimers, particularly where the data is licensed to many licensees under a form agreement or where the licensor is not in the business of licensing the specific type of data:
Carefully consider whether, given the nature and anticipated uses of the data, the disclaimers are acceptable. If the licensor resists a requested warranty on the theory that the licensor’s data is what it is, and has not been scrubbed, consider adding a knowledge or materiality qualifier.
It goes without saying that the licensor cannot grant the licensee broader rights in the data than the licensor possesses. So, it is important for the licensee to satisfy itself through due diligence and to document in the license agreement that the licensor possesses and is able to grant the licensee all of the rights the licensee requires to use the data for the anticipated purposes. This is especially true with respect to personal data where, in many cases, the licensor is not obtaining the personal data directly from the individual data subject. If notice to or the consent of the individual data subject is required, it is important that the licensor represents and warrants that it gave such notice or obtained such consent or that it obtained adequate assurances that the entity providing the data did so. In some cases, the parties will also need a mechanism that makes licensees aware if individual data subjects withdraw consent.
Finally, it is important to define when your rights with respect to the data begin and end. Often, data is licensed for a limited subscription term, with the understanding that it will be returned or destroyed at the end of the subscription term. However, for practical reasons, the licensee may require a perpetual license for data previously received and incorporated in the licensee’s systems. Given the proliferation of corporate databases and the ease with which data moves from one to another, it may be difficult or even impossible to track down the data. In addition, to the extent the data has been co-mingled with other data sets, it may not be feasible for the licensee to extract or stop using the data. Finally, many companies, such as financial institutions, will require a perpetual license to meet regulatory or control obligations to maintain the underlying data for decisions and actions.
This article originally appeared in the Mayer Brown e-book, Technology Transactions: Thriving in an Age of Digital Transformation . It is republished here with permission.
