Fintech 2023

Celebrating Five Years of FinTech Practice Guides with Reflections on the Future

It is with much appreciation for the team at Chambers and Partners that I write this essay, which will introduce the 2023 FinTech Practice Guide. This essay marks my fifth as contributing editor. When Chambers first approached with the idea of a FinTech Practice Guide in late 2018, I was excited. I had considered myself a financial services and technology lawyer for quite a while, but the term “fintech” was a much more recent name for what had been happening in financial services for a long time. To my mind, this new practice guide was catching the wave just as the term was gaining currency among policymakers, regulators, and investment bankers.

Against this backdrop, Chambers recruited quite a few excellent lawyers around the globe to discuss fintech law and regulation using an outline designed to show off the breadth and scope of the practice area for the first guide published in spring 2019. For the first introductory essay, I tried to do justice to the wide-ranging practice of fintech law by defining the term and talking about the trends and considerations that lawyers in the space faced every day. Chambers have kindly reprinted that original essay, as well as all of its kin, for this edition so you can read each of them in their entirety. The expansive definition of fintech advanced therein holds up reasonably well, as do the themes enumerated about the coming growth of blockchain as a technology, the need to return to first principles for effective regulation, the role of social media, and the necessity of both hard work and capital in order to build something truly revolutionary.

The 2020 essay, the second of this series, made a foray into thinking about whether the type of technology (named the “technology wrapper”) should dictate the law and regulation (the “legal wrapper” and “regulatory wrapper”) of the assets and items created or represented with that technology. The overall conclusion, favouring a technology-neutral approach, stressed the need for careful and considered analysis by regulators and lawyers in fintech so that first principles do not get lost. This idea should be more in demand now as policymakers and regulators grapple with some of the technologies I will discuss below.

In 2021, the introductory essay was a call to action around financial privacy. It argued that an individual’s financial information, such as the contents of their portfolio and transaction data, were just as much a financial asset as any stock, bond or fund interest in the portfolio, and bemoaned the fact that law and regulation failed to recognise this essential fact. The result of this failure has been an encroachment on personal freedom and an expansion of surveillance by both the public and private sectors. The tide has yet to turn here in any jurisdiction. It may never turn, because technology keeps delivering greater and greater accessibility of data. But I hope things change and financial privacy rights become enshrined in the pantheon of first principles.

Last year, the focus of the 2022 essay turned to token classification with the introduction of a “sensible” system. One might easily read it as an extension and refinement of the 2020 essay, and you can certainly see the beginning of many of the ideas for the sensible classification system when the 2020 essay peels back the technology wrappers to focus on what is really going on as the basis for analysis and application of first principles. I continue to strongly believe in the sensible token classification system and advocate for it wherever I go. Please join me!

What the Next Five Years Might Bring in Fintech

In rereading each of the prior missives, the backdrop is clear: the world of technology is ever-changing. We lawyers must face and embrace this change if we are to keep up, because technology, and not just financial technology, keeps revolutionising human activities. For this Guide overview, I decided to highlight several trends that I expect will occupy the fintech lawyer’s time in the next five years. We can check back in five years and see if I’m right!

Autonomously functioning code

Machine learning and smart contracts are both examples of autonomously functioning code. Once the code is deployed, it just functions. Although the code requires a prompt to start its actions, once that input occurs, the code continues to operate without human interference or direction. The output is often beyond the control of its creators, even if that output is completely as intended. To have a program work without human interference is the point. In this way, all who interact with it know the output or results of that interaction. Certainty and consistency are two of humanity’s greatest desires.

Of course, computer code is generally designed to function and create outputs. As I type this sentence, I am issuing prompts that the word processing software is interpreting as letters and punctuation, making them appear on the virtual page and in the machine’s memory. When I use a smart contract or the auto-pilot function on my car, however, there is not a constant series of human inputs that dictate everything the code produces. They take the initial input and deliver a (potentially) complex set of outputs and actions (driving a car is pretty complicated, but so is yield farming across the DeFi (decentralised finance) multiverse). More importantly, and especially so with smart contracts, once the process is started, it continues without the ability for a human to stop it. The inability to stop autonomously functioning code is perhaps the most novel element.

With more advanced machine learning and artificial intelligence, the code actually learns and “gets better” at its assigned tasks, even to the point of adjusting itself to do a better job. This self-optimisation, combined with autonomous action, holds many interesting possibilities for the future of robotics but also financial services. Will your next investment advisor be a self-optimising piece of software that goes out into the investing world on your behalf, doing trades, making investments, adjusting your portfolio and generally just doing things?

Questions abound about how the law should treat such software, and how it should be regulated. The traditional view that “oh well, it is just software and it might not work” has led to allowing broad disclaimers of liability and little legal responsibility for the creators or operators. Those ideas seem outdated in certain contexts. For example, we have an instinctive reaction that autonomously acting software in a car should not cause a crash, so the car company should be liable if it does. The same would hold true if the software investment advisor loses your life savings. We will need to start applying these reactions in other contexts, which will drive concepts behind new regulatory licensing, liability standards, and disclosure requirements. Perhaps even placing licensing requirements on code itself. Oh! brave new world and the software in it.

Open source code

We move from code that functions by itself, to code that is freely available for anyone to use. Open source is a generic term for software that is made available pursuant to one of a family of licences that more or less allow anyone to use the software for any purpose, including incorporating the code into other programs. The exact terms of the licences vary somewhat, with some requiring attribution or that any code into which the software is incorporated also be open source as well as other requirements. The key, however, is that the code is free for anyone to use without threat of copyright infringement and that the creators bear no liability for the code or its uses.

The implications of free code are many. For example, people worry about the business model that keeps the creation of such code alive as well as the ability of hackers and other bad actors to review the code, find the flaws and exploit them. There is also concern about ownership of any subsequently created code that uses even fragments of the original open source software. Finally, intellectual property lawyers also worry about the implications for patent and trade mark law when the code is part of another code base.

On the flip side, open source software has arguably created a revolution in programming with lots of freely available code that people can leverage to create new and interesting work. For example, it has fuelled the growth and development of blockchain and related ecosystems, including the Lego-like composability that DeFi proponents use to build increasingly intricate means of linking trading, lending and swapping protocols to achieve multi-faceted opportunities for transacting, perhaps previewing our autonomous investment advisor.

We cannot expect a fully open source world. The trend towards software that anyone, anywhere can use does seem irreversible precisely because it allows for all kinds of collaboration that is not present when the software is secret. This collaboration has the ability to expand what software can do, how humans can use it, and who has access to basic tools for building truly revolutionary products. The experimentation just in the blockchain sphere is proving how much will be built, but also the need for caution and scepticism. Hackers do exploit the flaws they find, business models are hard to create (although crowdfunding through blockchain token economies are changing that), and uncertainties about ownership for businesses may stunt adoption.

Living in this open source world poses challenges and delivers rewards. The direction of travel will depend on which outweighs the other. Lawyers will have many interesting issues to grapple with, not the least of which will be liability questions from the uses made of open source software. With open source, all the virtual world is indeed a stage and software are merely players, performers and portrayers to be arranged and rearranged at the writer’s whim.

Is gaming the testing ground for the future?

Games are for kids. Except when adults have some time to kill. Gaming contributes nothing to the real world. Except for the USD45 billion in revenue last year earned by gaming companies. Gaming is an escape from the real world. Except when incessantly driving progress in software and hardware.

Gaming can be thought of as a simulation for real life. Concepts can be tested in games without real-world consequences. Ideas can be advanced in games without changing reality. In-game activity is like a free opportunity to try out something new. It might work or it might not, but the cost is usually minimal.

Gaming universes have their own laws and rules, sometimes quite complex ones. Perhaps gaming gives us a chance to try out laws and regulations before they are implemented in a world as complex as our own. There remain untapped opportunities with gaming and simulation for policymakers and regulators but also for the financial services industry itself. I see our autonomous investment advisor running around again!

AI is coming for us all

And what of artificial intelligence? It is now possible to play around with fairly sophisticated machine learning models that regularly deliver wrong answers but give a startling likeness of a human response. They also deliver surprisingly correct answers, code, mathematical equations, and draft agreements. And that draft agreement can be done in the style of Tolstoy or Sophocles, Faulkner or Descartes, Martin Luther or Martin Luther King, Jr. Or, I suppose, your favourite partner at your favourite law firm.

This essay is written by a human with no AI intervention. But there is no reason that must be the case. I could have fed everything into the AI and asked for a draft, which I could then edit. The changes in lawyering wrought by Westlaw and Lexis-Nexis might seem quaint in a few years.

From a fintech standpoint, making AI available to investors need not mean an autonomous investment advisor. It can simply be an aid who gets you started in your investment journey, learning about your risk tolerance, investment goals and time horizons to create suggestions for your consideration. Why not do a side-by-side comparison to see if the AI portfolio does better than one you construct? Or your human investment advisor?

The possibilities seem endless, both exciting and a bit frightening. I am not keen to give up control, even over things I know someone else (or a machine) can do better. How the law and regulation will navigate these waters remains anyone’s guess. But there is no doubt that disclaimers and limitations of liability will abound unless we change that future.

Conclusion

We humans seem increasingly on the cusp of major technological changes in all areas of life. The MIT Technology Review every year publishes its list of impactful technologies. For 2023, most were hardware, including life sciences products, rather than software. If software will eat the world, as Marc Andreessen famously postulated, the law needs to adjust. It has so far been slow to do so, probably because we humans do not have good schematics for thinking about these issues. Wait a minute while I feed this question into ChatGPT.