Transactions involving fintech companies, and particularly fintech companies incorporating artificial intelligence ("AI") into their products and services, are now commonplace in the fintech landscape. CB Insights reports that AI startups are emerging at record rates, with 1,800 new startups raising equity for the first time since 2016, $19 billion of equity funding in 2018 and more than 5,000 equity deals across multiple industries since 20131. Legal and business transaction leaders should carefully consider the range of possible investments in companies offering AI products and services relating to financial services ("AI fintech companies"), and the potential risk and rewards of these investments.
For purposes of this article, we reference a Deloitte definition of artificial intelligence as "the theory and development of computer systems able to perform tasks that normally require human intelligence."2 AI has the potential (or likelihood) to transform the provision of financial services. Large financial institutions have traditionally been hampered by their legacy technology systems and cumbersome physical operations as well as the need to comply with complex and evolving regulatory requirements. As a result, a consistent theme is that incumbent financial institutions will need to collaborate with their AI fintech company disrupters, using commercial arrangements, partnerships and acquisitions to remain competitive. Incumbent financial institutions have advantages of their own, including large financial resources, the massive ability to manufacture compliant financial products, a wealth of data about their customers' financial activities and the deep-seated trust of their customer base, including tech-savvy millennials. Given the increasing speed at which AI and fintech are developing, the older sourcing strategies of "build versus buy" are being replaced with strategies that allow for flexible and rapid collaboration across a variety of licensing and capabilities acquisition models.
In this article, we will review the spectrum of possible AI investments -- ranging from licensing and service agreements to platform collaborations to financing transactions to joint ventures and strategic partnerships to minority and majority investments and, finally, to M&A-style acquisitions. We will also outline some of the due diligence, structure and contractual considerations for each type of transaction. We will focus on these considerations from the point of view of the buyer of, the investor in, the customer of or the lender to an AI fintech company, with potential AI fintech company counterparties including AI software licensors, cloud-based AI providers, financial data and analytics companies, and AI fintech platform companies. As described in this article, along this spectrum the financial institution may license AI technology, enter into an AI technology services agreement, enter into a "powered by" or white label commercial agreement, provide financing to or purchase whole loan assets of the AI fintech company, purchase a minority stake in or joint venture with the AI fintech company or acquire a majority interest in or all of the AI fintech company in an M&A transaction.
AI Licenses and Service Agreements
Licensing AI capability from an AI fintech company through a license or service arrangement is likely the fastest way to obtain AI for use by financial institutions. This may take the form of an on-premises license of AI that will be installed, trained and operated by the financial institution, or it may be offered as a "Software as a Service" solution in the cloud by the provider.
Many financial institutions are turning to a collection of AI fintech providers to test the waters. A good, lower risk way to do this is through a "proof of concept" arrangement. A proof of concept arrangement is a short-term agreement that allows a financial institution to test, and an AI fintech company supplier to prove, the value of an AI product or service.
Once the proof of concept is complete, the financial institution may license the AI from an AI fintech provider. Financial institutions should seek to satisfy the usual requirements for critical third-party service provider agreements in their AI licenses and services agreements. AI licenses present a few unique topics, including legal compliance of the AI decisions, allocating ownership and use rights of the components of AI, data use and privacy, and protection of intellectual property rights.
Legal Compliance. First and foremost, AI-based decisions must satisfy the laws and regulations that apply to financial services. This requires the financial institution to apply the same level of diligence to the AI tool or service that the financial institution applies to its other critical third-party products and services. Of particular concern is that AI-based decisions may discriminate because they rely on data that reflects a discriminatory past or looks only at correlation instead of causal factors. Financial institutions that use AI tools in credit decisions or fraud detection, for example, must ensure that these tools do not discriminate against certain protected classes of applicants or employees. AI tools used for insurance decisions will have to follow recently issued requirements from the New York Department of Financial Services on the use of "unconventional sources or types of external data" to address the risk of unlawful discrimination and a lack of data transparency.
In addition, AI systems should produce output that is transparent, auditable and that can be explained – sometimes called "Explainable AI." Licenses from an AI fintech company should address the extent to which the AI decisions and outcomes are explainable, and the method by which the financial institution may access those explanations and related data. The license agreement may also need to specify that the AI may be subject to regulatory examination, and require the AI fintech provider to cooperate with such examinations. Financial institutions may also want to require that AI has "circuit breakers" – a method for pausing operations to gather data about correct and compliant operation, confirm security compliance, and make necessary adjustments in the AI tool to eliminate errors, mistakes and bias. Record-keeping and audit requirements are also important considerations for financial institutions. Because AI tools evolve, data sets change and iterations are part of the process, financial institutions should address how they can access versions of past decisions based on AI tools and data sets that have shifted over time. This is particularly important when financial institutions are using AI in a provider cloud and when the financial institution is not in control of archiving the AI components and outputs.
Financial institutions can mitigate these AI risks by utilizing oversight, risk management and controls to meet legal compliance and business objectives, and by incorporating provisions addressing these requirements in the AI license. Finally, consider whether financial institutions should include rights to training and access to specialists who are familiar with the AI tools and can assist the financial institution with its training, use and ongoing monitoring requirements. Regular compliance meetings with the provider may be required to provide assurance on these key items.
Allocating Ownership and Use Rights and Training Obligations. There are a number of important questions for financial institutions to consider regarding contractual ownership and use of the components of AI in their licensing agreements. These components include the AI tool, evolutionary changes to the AI tool, the training data and instructions, and the output of operation of the AI tool. When licensing AI, most AI fintech providers will expect to continue to own the underlying AI tool, and some may expect to own the evolutionary changes as well. Much of the AI that financial institutions will use may require training. The license should address which party will train the AI, which party will own the training instructions, and which party will own the evolutionary works of the AI tool based on the training. Shifting to the output of the AI tool, most financial institutions would expect to own the decisions and the decision criteria of the AI tool, and this must be specified in the license agreement to achieve that result. Once the parties have determined how they will allocate these ownership rights, they also need to determine whether and to what extent the other party will have ongoing license and use rights in those components.
Data Use and Privacy. Data is the fuel for AI, but data use must comply with the privacy, data security, export control and other laws that apply to the data. In addition, data use must comply with any contractual requirements to third-party data suppliers. These are often not well understood. To guard against these data pitfalls, financial institutions should inquire as to the level of legal and regulatory diligence that has been done on the uses of data to fuel AI systems. The license should specify whether the AI will rely on provider data or financial institution data or both, and, importantly, which party will own which data, and which party may use that data and for what purposes. The license agreement may also specify that the party supplying the data is responsible for obtaining necessary consents and rights to use that data for the AI, and address liability for issues arising from improper use or failure to obtain proper consents. If financial institution data are used for the AI, and those data include non-public personal data, the financial institution will have to assess compliance with its privacy policies governing that data. Similarly, many countries, such as European countries, have tough data protection laws that prohibit the use of individual data for automated processing to evaluate any feature of behavior, preferences or location absent the explicit consent of the individual, and yet, automated processing of individual data to determine preferences is the hallmark of many AI tools. Consider whether the license should require the provider to conduct privacy assessments of the AI tool on a periodic basis.
Protection of IP Rights. Patent, copyright, trade secret and other IP laws were written with a bias to protecting human creativity. Intellectual property (IP) laws in the United States do not square nicely with AI. Not only may a financial institution not own AI that it pays to create, it also may not have the means to fully protect its AI under U.S. IP laws. Contractual protections are a key element of capturing and preserving value in the creation of and returns on the investment in AI. These protections, to be effective, must be implemented before the AI effort begins, and will rely on clear statements of ownership and use rights in the various components of AI as addressed above.
Service agreements in which AI fintech providers use or rely on AI on an incidental basis to deliver the services are another channel through which financial institutions may obtain the use of AI. Although the AI may not be the cornerstone of such an arrangement, financial institutions should require service providers to reveal if they are using AI tools to provide the services, and if so, they should understand the uses. If the uses bear on any of the issues described in the prior paragraphs above, then the financial institution should take care to perform diligence on those uses, and to define the contractual rights and obligations with respect to such AI as part of the service agreement.
Platform Collaboration and White Label Arrangements
Broadly defined, a digital platform is an integrated framework of digital tools and services that implements key business processes to facilitate exchanges between producers/manufacturers of services/products and consumers. Put more simply, it is the foundation on which a digital business is built. The difference between digital platforms and previous methods of technological transformation is the exchange function of a platform. In addition, platforms are not simply cost-saving technology for companies' back-office functions; instead, digital platform technology is transforming customer-facing, revenue-generating functions.
These exchanges vary in terms of openness and complexity. While one often thinks of platforms as vast "many-to-many" systems (e.g., Facebook, Google, AirBnB, Uber), a platform can also include more traditional exchanges where a single producer is trying to reach many consumers. These traditional exchanges are exemplified by fintech platforms, which can include systems for consumer banking (e.g, SoFi, Stash), retail investment (Robinhood), payments (Venmo, PayPal, Zelle), loan origination (Lending Tree and multiple white label lending platforms for individual banks), and financial advice (Robo-advisors, H&R Block, Watson).
A financial institution may seek to enter into a commercial arrangement to white label an AI fintech company's digital lending or other digital financial services platform for use by the financial institution. These commercial arrangements – known as a "platform collaboration" or a "white label" or "powered by" arrangement – allow the financial institution to obtain AI capabilities as opposed to building its own AI capabilities. Some of the advantages of a platform collaboration include: (1) relatively small investment for the financial institution; (2) the financial institution gains efficiency because it is not "reinventing the wheel" where AI solutions may already exist in other formats; (3) faster time to enter the market because developing AI is outside of the financial institution's core competency; (4) the financial institution can focus on its core competency; and (5) platform or white label arrangements allow for scalability.
There are, however, risks and disadvantages that must be addressed in any platform collaboration. As noted in the "AI Licenses and Service Agreements" section above, data security and privacy are major issues that the financial institution must consider carefully. Further, under this type of arrangement, the financial institution may have very little control over the direction of the AI platform. Lastly, AI fintech company providers are often time-hungry, highly leveraged start-ups seeking to maximize the rapid growth that follows from successful early entry into an AI fintech company space. Thus, the financial institution must consider the financial stability of the AI fintech company provider and include adequate protections in the contract (e.g., termination rights for financial degradation, rights to retrieve data in a usable format upon request and termination assistance rights to facilitate a smooth transition to an alternative platform).
In negotiating a platform collaboration contract or a white label arrangement for AI capabilities, a financial institution may find it helpful to leverage its experience from contracting with providers of ERP, information technology (IT) infrastructure and other back-end technology services. In fact, many of the concerns described above in the "AI Licenses and Service Agreements" section are also present in platform collaboration and white label arrangements.
Even the most experienced financial institutions, however, will face unique issues when it comes to platform collaboration deals or white label arrangements for AI capabilities. One such issue is legal compliance. Similar to licensing agreements for AI capabilities, the financial institution must ensure that the white label services and the financial services platform (including the AI tool) comply with all laws and regulations that apply to financial services. The AI fintech provider will most likely try to limit its obligations to complying with laws applicable to the AI fintech provider in its provision of the services. That universe of laws is generally small, and the financial institution may seek to allocate more responsibility on the AI fintech provider for violations of laws applicable to the financial institution that are caused by the AI fintech provider. The parties will need to find a middle-of-the-road approach that provides adequate protection for the financial institution. One compromise for the financial institution to consider is to require the AI fintech provider to bear responsibility for (a) complying with laws applicable to the AI fintech provider in its provision of the services and (b) violations of other laws caused by the AI fintech provider's failure to follow the financial institution's written instructions with respect to such other laws. Another compromise is to require the AI fintech provider to bear responsibility for complying with (x) laws applicable to the AI fintech provider in its provision of the services and (y) any laws that are applicable to the financial institution (but not to the AI fintech provider as a technology provider of the services) provided that the financial institution informs the AI fintech provider of such laws in advance.
Another thorny issue in platform collaboration contracts and white label services arrangements is the ownership rights for developed IP. As mentioned in the "AI Licensing and Service Agreements" section, the parties need to clearly allocate IP rights. The parties need to consider who will own the developed IP that incorporates both the financial institution's and the AI fintech provider's proprietary materials. For example, the developed IP may combine fraud models from the AI fintech provider and underwriting criteria and credit policies from the financial institution. If there are practical challenges in separating that combined, developed IP upon termination of the contract, the parties may consider requirements to delete or destroy that IP upon termination. The parties, however, will need to assess this issue on a case-by-case basis, depending on the circumstances of the deal.
Financing AI Fintech Companies
There are a variety of financing options available for financial institutions lending to, or investing in, AI fintech companies. The type of financing that the lender will execute typically relates to the AI fintech company's experience in the finance industry as well as the space in which the AI fintech company wants to brand itself – technology or finance. Assuming that the AI fintech company's business model is to make loans to customers, most AI fintech start-ups and AI fintech companies without extensive experience in the financial services industry enter into whole loan sale transactions with various investors or lenders before moving onto capital markets transactions. The motivation for the AI fintech company is threefold, as these types of transactions: (1) allow AI fintech companies a flexible relationship with an investor or lender memorialized in a few documents that can easily be amended and do not trigger significant regulatory compliance, (2) provide exposure for the AI fintech company to various investors and lenders and (3) are structured as off-balance sheet for accounting purposes.
From the point of view of both the financial institution acting as the investor or lender and the AI fintech company, whole loan sales with a single investor or lender are not structurally complex transactions that trigger extensive regulatory compliance and diligence. Instead, these transactions are usually structured as a one-time (or multiple, scheduled) sale(s) from the AI fintech company to its investor or lender where the AI fintech company and the investor or lender agree to the sale(s) on certain negotiated terms. Additionally, the AI fintech company agrees to service the assets and undertakes the servicing responsibilities in the transaction documents. Given the nature of AI fintech companies, servicing is a crucial component for the investor or lender to consider in financing transactions. Servicing responsibilities usually include collecting payments from the underlying obligors on the assets, monitoring the activity of the underlying obligors, enforcing the obligor contracts, taking action to maximize collections in the event of obligor delinquency or default, and providing the requested servicing and performance data to the investor or lender. While the AI fintech company does need to comply with its general corporate and licensing regulatory requirements, this structure does not trigger the typical Dodd-Frank regulatory requirements or generally require registration with the Securities and Exchange Commission. Finally, since there are not multiple transaction parties, the AI fintech company and investor or lender can more easily amend the deal documents if changes are needed as the AI fintech company hones its data systems and servicing policies and procedures.
Since whole loan sales can be papered by a handful of documents, an AI fintech company is able to easily enter into multiple transactions with various investors or lenders. By having exposure to various investors and lenders in a whole loan sale program, the AI fintech company accesses liquidity from multiple sources, which also lowers the financing risks for any single investor or lender. Additionally, whole loan sale investors may be non-bank private equity or hedge funds that often seek leverage from larger, more traditional financial institutions, providing exposure for the AI fintech company to financial institutions that the AI fintech company may not be able to obtain on its own. Accordingly, whole loan sales set the stage for more complex financing transactions in the future.
Finally, investors and lenders typically structure whole loan sales as a true sale from the AI fintech company to the third-party investor or lender. This type of transaction is appealing both to the AI fintech company, since it allows it to obtain financing while easily achieving off-balance sheet treatment through a true sale to an unaffiliated third-party investor or lender, and to the investor or lender, since it should provide isolation from bankruptcy risk. By achieving off-balance sheet treatment, AI fintech companies are also more easily able to brand themselves as technology companies rather than companies that operate in the financial services space.
Investors and lenders may also offer their AI fintech companies financing through a warehouse facility. A warehouse facility is typically negotiated between the AI fintech company and an agent bank lender. These types of facilities are often syndicated to a group of investors or lenders through the agent. Additionally, whole loan investors and lenders will often provide financing of the equity piece under these structures. Warehouse transactions provide investors and lenders with another option to finance AI fintech companies that is slightly more complex than a whole loan sale, but not as sophisticated and regulatory intensive as a capital markets transaction.
While whole loan sales and warehouse loans offer AI fintech companies relatively straightforward access to liquidity from a variety of financing sources without necessitating significant regulatory compliance, it is, nonetheless, advantageous to maximize funding options through a variety of finance transactions. Investors and lenders providing whole loan and warehouse loan facilities will seek the ability to take out their financing through capital markets transactions. While some financial institution investors and lenders may be comfortable purchasing whole loans, others may prefer to purchase securities backed by such loans for risk and liquidity purposes. Thus, in addition to whole loan sales, AI fintech companies may look to access the capital markets and, more specifically, the structured finance markets. While securitization transactions can provide a more efficient cost of funds for the AI fintech company, investors and lenders will require attention to significant additional regulatory requirements and the AI fintech company will need to have adequate legal, compliance, systems and servicing procedures in place to provide the data and access to employees necessary to facilitate compliance. The financial institution acting as investor or lender may also act as underwriter, initial purchaser or placement agent for the securitization. The underwriter will assist the AI fintech company entering into a securitization, which typically requires the following:
- static pool data on prior transactions or vintage data and pool data relevant to the assets included in the transaction;
- customary narrative descriptions of the company's material underwriting and servicing practices, and other written information for use in an offering document, such as disclosure on the legal and business risks relating to AI-based products and services;
- holding 5% of risk in the transaction;
- coordination with accountants to facilitate the provision of a customary agreed upon procedures letter by an independent accounting firm;
- allowing reasonable access for rating agencies and the investment banking firms to the company's origination and servicing personnel and its records relating to the assets to be securitized and employees with responsibility and knowledge with respect to the securitized assets;
- maintenance of a 17g-5 website allowing any nationally recognized rating agency to access information about the transaction;
- undertaking to make certain filings with the Securities and Exchange Commission; and
- additional requirements if the securitization will be a public offering of securities.
As AI fintech companies enter into financing arrangements with third-party investors or lenders, both AI fintech companies and the investors or lenders should consider the different funding options available to a growing AI fintech company. While whole loan sales provide access to liquidity without as many extensive or complex diligence and legal requirements, not all investors and lenders want to hold loans, and warehouse facilities and capital markets transactions typically have a higher dollar amount. Conversely, while warehouse facilities and capital markets transactions require more diligence and regulatory compliance, they offer access to high dollar bond issuances with multiple sophisticated third parties. AI fintech companies without extensive experience in the financial services industry, as well as their investors and lenders, should consider these factors when establishing funding plans.
Joint Ventures and Strategic Partnerships
The term "joint venture" is quite broad and can involve creating a new entity, an ongoing contractual relationship or a combination of both. As distinguished from a strategic investment or an M&A transaction, a joint venture typically involves two or more parties that come together to achieve a common goal for profit.
In the current regulatory environment, it may be relatively rare for large financial institutions to joint venture or partner with an AI fintech company in the traditional sense, but other large non-bank finance companies may consider the joint venture structure attractive. As discussed below, a large financial institution, such as a bank holding company or an insurance company, is typically highly regulated and seeks to avoid obtaining "control" of the AI fintech company, in most cases by keeping a minority equity investment below 5% (or 10% in the case of an insurance company) of the AI fintech company's voting shares and otherwise avoiding indicia of control. Indicia of control include holding a voting seat on the company's board of directors, certain veto or consent rights, entering into a management agreement or entering into significant business or commercial relationships with the AI fintech company. If the financial institution seeks a control relationship, it may be simpler to acquire complete control through an acquisition as opposed to a joint venture or partnership. On the other hand, the financial institution may forego any equity investment in order to avoid these control questions and seek only a commercial or financing arrangement as discussed above.
Assuming that the joint venture partners are willing to have their joint venture entity be treated as a regulated entity or the joint venture entity is otherwise not subject to what may be viewed as burdensome bank or insurance regulations, there can be a number of advantages to using a joint venture entity as opposed to a contractual joint venture. These advantages include: (a) access to technology, subject matter experts like data scientists, and products contributed to the joint venture as well as distribution channels and markets with greater economies of scale; (b) sharing of regulatory risks that accompany financial institutions, especially when entering a new market; (c) internal and external constituencies (e.g., employee talent in the joint venture and end users of the technology) will perceive a separately identifiable and visible enterprise conducting the joint venture business, with the venture lending itself more to AI innovation than to regulated bank or insurance activity; (d) interests in a joint venture are generally easier to sell or transfer than a collection of contractual relationships; (e) the joint venture entity creates an independent vehicle with greater flexibility and convenience for capital-raising activities; (f) the joint venture entity provides a familiar structure (e.g., a corporation, limited liability company or limited partnership) in which management and governance rules can be established and in which directors, officers and employees typically play familiar roles in making decisions and implementing them, with this level of oversight likely being important in the developing area of AI; (g) the joint venture entity provides a convenient vehicle for measuring profits and allocating and distributing them to the joint venture parties; (h) the joint venture entity can be an independent employer providing identification and focus for employees, including incentive compensation such as equity interests and the opportunity to work on cutting-edge AI projects; (i) the joint venture entity largely enables the joint venture parents to insulate themselves from the liabilities of the joint venture business; and (j) the joint venture entity creates the potential for flexibility in addressing tax matters.
Disadvantages of the joint venture structure – other than the perhaps overriding concern that the AI fintech joint venture will become a regulated entity based on its control by a regulated financial institution – include: (a) complexity because establishing a separate joint venture entity often involves initial and ongoing issues, tasks and costs that are not necessarily present in a contractual association, with time-consuming oversight required by senior managers of the alliance participants; (b) a likely more complicated unwind process because assets, contracts, employees and other resources of the joint venture business may be property of, or affiliated with, the joint venture entity; (c) loss of control in that the joint venture business will normally be, in large part, conducted by the joint venture entity and the rights and ability of the joint venture entity and its activities will be limited by the governance rules of the joint venture entity; (d) difficult fiduciary duty and conflict of interest issues may arise with a joint venture entity that may not arise in a contractual joint venture (although these can largely be handled contractually); and (e) the contractual joint venture can allow more flexibility in staging and developing the joint venture by establishing an initial "let's get our feet wet" relationship without the more substantial commitment involved in establishing, and providing assets and other resources to, a separate joint venture.
Stock Investments and M&A Transactions
Strategic investments and M&A transactions offer a large financial institution, such as a bank or insurance company, some additional flexibility to tailor an investment to its specific business strategy, with each structure having its own unique advantages and disadvantages. Two general concerns applicable to each structure are: (1) the "control" analysis described above in the "Joint Ventures and Strategic Partnerships" section and the effect of bank or insurance regulatory control on the AI fintech company; and (2) the level of diligence a potential investor should complete with respect to each structure. In this section, "investor" refers to financial institutions as investors in or acquirers of AI fintech companies.
A passive, non-controlling investment can offer a large financial institution investor and the AI fintech company a number of advantages. These advantages include: (a) allowing the investor to leverage the AI offerings of the AI fintech company in its business with relatively low risk to the investor due to a limited commitment of resources; (b) potentially less stringent due diligence requirements of the AI fintech company, in general, than majority investments and M&A transactions, but this can vary depending on the cost/benefit analysis and risk tolerance of each individual investor; (c) the imposition of fewer regulatory burdens on the AI fintech company; (d) allowing the AI fintech company to leverage the infrastructure and expertise of the investor; and (e) the AI fintech company's retention of a certain level of autonomy. Disadvantages of this structure include: (w) very limited investor control over the AI fintech company's activities (e.g., no board seat, very few consent rights over activities of the AI fintech company, etc.); (x) limited investor protective provisions; (y) requiring the investor to conduct a relatively complex and ongoing control analysis for regulatory purposes; and (z) tension created due to the differing goals of the investor (financial return) and the AI fintech company (long-term viability). The obligation of the investor to continually assess its level of control over the AI fintech company to avoid subjecting the AI fintech company to regulatory oversight is a key disadvantage to a minority investment. For example, a bank holding company investor must ensure its equity investment remains below 5% in addition to monitoring other means of exercising control over the AI fintech company, such as the appointment of a board member, veto rights over certain actions of the AI fintech company, ownership of 25% or more of any class of voting securities, rights of first refusal and ownership of convertible securities.3 As a protective measure, a minority bank holding company investor should seek to include certain transfer rights, such as a put right, for itself in connection with its investment to allow the investor to exit the AI fintech company if regulatory concerns arise.
Investments by insurance companies (or their affiliates) will potentially be subject to the laws governing insurance holding company systems in the states where the insurance companies are domiciled (or deemed commercially domiciled). Generally, those laws presume control – and thus an affiliate relationship – to exist where one person, directly or indirectly, owns 10% or more of the voting securities of another person, although that presumption can be rebutted by submitting a disclaimer of control to the domiciliary state insurance commissioner. In addition, other types of rights, such as the appointment of board members, may be deemed by an insurance commissioner to constitute control of an entity. The laws in many states limit the ability of an insurance company to acquire a controlling minority interest in another entity. In addition, if an entity is treated for insurance regulatory purposes as an affiliate of an insurance company, that relationship will need to be disclosed in the insurance company's statutory financial statements, annual holding company registration statements and enterprise risk reports, and the domiciliary state insurance commissioner will need to be notified in advance of material transactions between the insurance company and its affiliate, giving the commissioner an opportunity to review the transaction before it can go into effect.
Alternatively, if a large financial institution seeks a control relationship, it can structure its investment as a majority investment or an M&A transaction. Some advantages of a majority investment include: (a) providing more investor control over the AI fintech company than in a minority investment; (b) allowing the investor the opportunity to enhance the operational efficiency of the AI fintech company and address any existing risks (e.g., amend existing material agreements to address deficiencies); and (c) providing the AI fintech company with a greater opportunity to leverage the infrastructure and expertise of the investor. Disadvantages of a majority investment include (w) subjecting the AI fintech company to regulatory oversight; (x) requiring a much larger resource commitment from the investor, which entails a higher level of risk, necessitating a much higher level of due diligence (query whether it may be more advantageous to acquire the entire AI fintech company); (y) requiring a higher level of investor responsibility and oversight with respect to the operations of the AI fintech company, including regulatory compliance; and (z) integration issues with respect to the cultures of the investor and AI fintech company. The effect of the investor obtaining control of the AI fintech company is one of the most important factors for the investor's consideration. Generally, majority investments require a much more thorough due diligence investigation of the company than minority investments. The investor will need to assess the AI fintech company's current operations and marketing strategies (including the AI fintech company's website) and review its contracts, in each case with a particular focus on data security and regulatory compliance, as discussed more fully below.
In extreme cases, it may be necessary to shut the AI fintech company down for a period of time to resolve any major issues identified in due diligence.
Lastly, a large financial institution may wish to acquire full ownership of an AI fintech company in an M&A transaction. Each of the advantages and disadvantages of a majority acquisition apply to an M&A transaction, often to a greater extent. A key additional advantage of an M&A transaction is the flexibility provided, more specifically the opportunity to utilize a number of different structures to address specific risks (e.g., the use of an asset sale to protect against pre-closing liabilities). Some key disadvantages of M&A transactions include (a) requiring the highest level of due diligence and (b) concerns related to retention of key employees are at their peak.
The buyer's due diligence of an AI fintech company in an M&A transaction should include a confirmation of ownership of intellectual property and software, a personnel assessment and an evaluation of regulatory and data privacy risks. Analyzing the source code underlying the IP is critical. Open source code licenses may require disclosure to the public domain of all or a portion of the source code into which the open source code subject to any such license was incorporated. To reduce its risk, the M&A buyer should also seek to negotiate strong seller representations in the transaction documents with respect to matters such as ownership of IP, outbound licenses of the IP, use of open source code, the formatting of the source code (i.e., that it has been documented in a manner that enables a programmer of reasonable competence to understand it, manipulate it, etc.), compliance with data protection laws and best practices, and other similar matters.
The buyer of an AI fintech company should also seek to address due diligence issues and risks that are particular to AI providers. For example, the buyer should include compliance with law representations and covenants that allocate strict liability to the seller for machine learning output regardless of whether any breach is "intentional" or "negligent" or is known by the seller. Particularly where the AI fintech company engages in lending or making underwriting decisions, the buyer should address liability for discrimination and fair lending compliance, including for any disparate impact. The buyer may also seek a representation that decisioning criteria are "explainable" or at least diligence the design criteria of the AI fintech company for explainability. Cybersecurity and data privacy representations and covenants may also need to be augmented in light of data-intensive AI systems.
The buyer may seek to impose covenants in an M&A transaction that obligate the AI fintech company to address certain issues prior to closing, such as requiring the AI fintech company to bring its operations into compliance with data protection laws (including implementing any necessary changes to its IT systems), engaging a consultant to undertake a review of open source code, making changes to its marketing materials, obtaining any additional state or third-party licenses to operate the business, or renegotiating or terminating certain problematic contracts. Depending on the M&A buyer's leverage, it should also consider including closing conditions related to these matters to avoid being forced to close the acquisition and make these changes itself post-closing, which shifts the risks associated with any necessary shutdown to the buyer.
Lastly, as part of its due diligence process, the M&A buyer should identify key employees to retain following the closing. As mentioned above, there may be substantial differences between the cultures of the financial institutions buyer and the AI fintech company. Employees will often be moving from a relatively autonomous position with modernized infrastructure at the AI fintech company to a much more structured environment, often with restrictive and outdated legacy infrastructure, at the buyer. Considering the importance of key employees, such as lead software engineers, to the AI fintech company, the buyer should ensure it is offering attractive compensation packages to encourage these employees to remain following the closing.
As shown in our discussion above, transactions involving investments in AI include a wide spectrum of possible structures, with legal and business issues that vary based on the transaction type. Financial institution investors should first define their AI goals and strategy, and then attempt to align their investment tactics with their AI strategy. As these AI strategies evolve, so will the transactions for investing in AI.
* Mses Baker, Eisner and Raymond and Mr. Pennell are partners at Mayer Brown LLP. The authors gratefully acknowledge the assistance of Corina Cercelaru, Lawrence R. Hamilton, Joshua La Vigne and Donald S. Waack in preparing this article.
1. CB Insights, What's next in AI?, www.cbinsights.com, page 10.
3. Note that a potential alternative path for a bank holding company that has elected "financial holding company" status to invest in AI fintech companies is under the merchant banking authority in section 4(k)(4)(H) of the Bank Holding Company Act. This article will not attempt to address merchant banking authority, in part because its requirements (including with respect to the "routine management or operation" of a merchant banking portfolio company) are relatively restrictive.
Originally published October 21 2019
Visit us at mayerbrown.com
Mayer Brown is a global legal services provider comprising legal practices that are separate entities (the "Mayer Brown Practices"). The Mayer Brown Practices are: Mayer Brown LLP and Mayer Brown Europe – Brussels LLP, both limited liability partnerships established in Illinois USA; Mayer Brown International LLP, a limited liability partnership incorporated in England and Wales (authorized and regulated by the Solicitors Regulation Authority and registered in England and Wales number OC 303359); Mayer Brown, a SELAS established in France; Mayer Brown JSM, a Hong Kong partnership and its associated entities in Asia; and Tauil & Chequer Advogados, a Brazilian law partnership with which Mayer Brown is associated. "Mayer Brown" and the Mayer Brown logo are the trademarks of the Mayer Brown Practices in their respective jurisdictions.
© Copyright 2019. The Mayer Brown Practices. All rights reserved.
This Mayer Brown article provides information and comments on legal issues and developments of interest. The foregoing is not a comprehensive treatment of the subject matter covered and is not intended to provide legal advice. Readers should seek specific legal advice before taking any action with respect to the matters discussed herein.