Displaying and Evidencing Contract Terms in a Post-Visual Era
Our current contracting processes rely on the human sense of vision. With limited exceptions, we use our eyes to read (or, far too often, not read) and agree to contract terms. At the same time, we now see a rapid growth of non-visual devices to interact with others using audio, sensors, smart contracts, artificial intelligence, and intelligent agents. Technological developments increasingly force us to reconsider core assumptions and approaches. How will we make contract terms available and accessible in understandable ways when contracting happens in a non-visual context? How will we evidence and document agreements without visual forms of contracts? Will our approaches change how we contract in the visual world? In this article, I consider these questions, given the history of contracting and the issues new technologies have raised, and review some promising current and future approaches to addressing problems posed by traditional contract delivery methods when they meet new and evolving technologies.
Brief History of Contract Delivery
In recent years, archaeologists have discovered new evidence of historic forms of contracting from Sumerian, Roman, and other civilizations. Evidence of transactions was captured on stone, clay, papyrus, wood, and other forms. Given the difficulty of using these materials, it’s no surprise that the contract terms are minimal. However, we see descriptions, pricing, signatures, and other contractual terms in their most elemental forms.
In historical context, multi-page agreements with complex terms are a relatively recent development in the history of human agreements. If we created a timeline of contracting, our current approach to recording contracts would be both relatively recent and only a small segment of the entire timeline. We would also see that other approaches have come and gone.
Our current “long-form” approach to contract display and evidencing required several technology advances and adoptions. Stone or clay tablets would not work. We needed a workable medium like inexpensive paper or digital screens. We also needed writing or document creation technologies to produce long and comprehensive contracts. In earlier times, we also needed enough humans to throw at the project. Think of the pictures of long rows of low-paid scriveners at desks handwriting documents in the nineteenth century. Monks in monasteries were working on capturing the knowledge of the world, not writing contracts.
In the twentieth century, typewriters and fast typists made it possible to create longer contracts with the benefit that the contracts were readable and legible as compared to handwritten contracts. However, long documents could take time to produce, often had errors, and even a simple revision might require retyping an entire page.
Word processing became the key technology that brought us to our current approach to contract production and display. Creating and printing long documents became an easy task. We could create and improve templates, reuse contract provisions, add terms, and become more complete and thorough in our approach than ever before. A five-page indemnification provision is no more difficult to add to every contract than a one-paragraph provision would be. The Internet gave us clickwrap and clickthrough agreements that are rarely printed, allowing the length and complexity of contracts to keep expanding.
Evidencing of contract agreement has also changed through history. In simple terms, we want to confirm that the person entering into the contract is who they say they are (identity or authentication), they have the authority to enter into the contract, and that we have evidence of the agreement and its specific terms.
These elements are not trivial, and it is fascinating how they have evolved. Seals, badges, authorization letters, symbols of identity, and other approaches that have been used throughout history as indicators of identity and trust. Technologies like writing instruments, inks, waxes, and, today, digital signatures have evolved. We have tried practices like signing, initialing every page, notarization, and requiring witnesses. In our digital world, we have tried to replicate those practices with varying degrees of success. For example, consider the difficulties witnessing and notarizing requirements caused in the earliest phase of the COVID-19 pandemic.
The adoption of the Uniform Electronic Transactions Act throughout the United States in the early 2000s made it easier to evidence contract signatures and acceptance by digital means, including non-visual means (e.g., a voicemail message).[1] The premise of these laws is that an electronic signature should be treated the same as a physical or “wet” signature. In other words, if you text me a “thumbs up” emoji in an instant message reply to my text “I’ll sell you my ebike for $750. Venmo me,” we very likely have a valid, signed contract.
There are many other examples of digital approaches to contract formation and signing. Clickwrap and clickthrough agreements on the Internet are a daily part of our lives, whether we read them or not. It is common to see statements like “Your use of this website affirms your agreement with our terms.”
History is also replete with challenges created by contracting processes. Literacy, language, and dialect differences can create problems. Tablets can break into pieces. Paper can be damaged by fire or water. Contracts printed by old thermal printers simply disappear. Typeset and online agreements are often presented as if they cannot be negotiated. “Fine print” is a pejorative term. It can be cumbersome to find witnesses and notaries. Many of our current contracts are not even read. To be fair, many of our contracting experiences (e.g., buying a car) seem like the equivalent of handshake deals because we do not read most of the terms and sign whatever is put in front of us.
Note, however, in all of these examples, there is a core assumption that terms and evidence of agreement require visual elements and the ability to see them (or to have the visual elements converted into a format that can be used by those who cannot see then).
Moving to Non-Visual Contracts
The legal profession and legal system’s current and growing challenge is addressing the rise of non-visual transactions, interfaces, devices, and their implications. What should we do with identity/authentication, authorization, and evidence of agreement with specific terms as interactions happen more often than ever on non-visual platforms?
By non-visual, I mean contracts without any visual elements. I will focus on four examples.
First, contracts when no screens or displays can be involved. In the world of the Internet of Things, we interact with devices and sensors without screens or displays where contract terms could be shown. Think of a variety of “smart” devices: smart lightbulbs, toothbrushes, or running shoes.
Second, contracts where all of the interaction is through voice and audio only. Examples include virtual personal assistants like Alexa and Siri, smart speakers, and other voice-controlled devices.
Third, contracts that, at least conceptually, happen at a code level, without human engagement. That brings us into the world of intelligent agents, smart contracts, and, perhaps, artificial intelligence.
Fourth, contracts that happen and continue to happen invisibly once you are on a platform. Examples would be seamless and invisible transactions on gaming platforms, the metaverse, shopping or trading platforms, and the like. There might be a visual component of what you use on the platform, but the contract terms are not apparent or visible. This area is already large. For example, eBay is currently handling more than sixty million dispute resolutions a year.
If we look at each of these examples in light of the identity/authentication, authorization, and evidence framework, the challenge each poses is readily apparent. Each challenge is significant. Each example also illustrates my initial point that our new technologies push us to face core and fundamental issues.
How do we communicate a long and complex contract when the only medium is audio? What does that tell us about what is essential and required in the contracting process? If people do not read clickthrough agreements now, why would we think that they would listen to them?
Do intelligent agents have authority to agree to terms and how do we determine that? What happens if a contract party sends out terms that cannot be displayed on a device used by the other party? How does a platform keep track of the many agreements with many parties that might take place on the platform?
Finally, how do we as individuals or businesses keep track of non-visual contracts, retrieve them, prove them, and resolve disputes about them?
In other words, how do we think not just about legal enforceability but the practical parts of the contract process when we move out of a visual process to either something that is non-visual or a hybrid?
Learning from the Past
If, as I believe, we are in a period of transition, we can look at other periods of transaction for learning and guidance. What we will see is that transition comes with experimentation.
Our most recent transition – from paper to digital over the Internet – is instructive, but other historical times carry their own lessons. None of the lessons learned were especially easy and most took time. We learned a lot at the edge cases where significant issues became apparent.
The work of Lisa Colpoys and others at Illinois Legal Aid Online over a period of 20 years is instructive.[2]They focused on the many issues involved in delivering legal help over the Internet, including language and literacy issues, barriers to and difficulties of technology access, levels of distrust, and much more. There was an evolution through experimentation with successes and failures, with hard-won lessons that can be applied in other contexts.
There is also much to be learned from the legal design movement. Tessa Manuello and Margaret Hagan are examples of leaders in this approach.[3] Although the focus is primarily on visual elements, the principles will help us address non-visual issues.
These issues are not uniquely legal or contracting issues. Other consumer product companies are working on these issues. If you have used an Amazon Echo device, you know how much it has been improved. How did you enter into contract terms with Alexa, what are they, and where are they kept?
As with so many things these days, the core issue might be trust. What approaches improve our level of trust and what ones diminish it? Blockchains were conceived as a form of providing “trust” in a decentralized way where there was insufficient trust in centralized authorities and approaches. Might blockchains play a role in non-visual contracting?
Applying What We Have Learned and Possible Approaches
As you read this article, people are working on using artificial intelligence to analyze contract terms to identify over-reaching and problematic terms. Does that mean that artificial intelligence will “solve” the challenges raised by non-visual contracts. That is not likely. Will AI be part of our approach? That is more likely.
Our approach to non-visual contracting will be experimental, layered, and draw on what we have learned in the past. We would also like it to be data-driven and evidence-based. Our best starting point might be the scientific method – form hypotheses, test them, and change our hypotheses based on results.
When I think about our current contract approaches, I have some observations that have led to the following hypotheses, all of which can be challenged and must be tested.
1. We have gradually moved to a place where only lawyers, not the parties, read and understand contract terms.
2. People are surprisingly comfortable with agreeing to contract terms they do not know or understand.
3. A large component of this comfort comes from the belief that some authority (e.g., a court or regulatory body) will not enforce over-reaching or unfair terms.
4. Another part of this comfort comes from a belief that the vast majority of contracts do not end in disputes.
5. People have come to believe that the terms of most consumer contracts cannot effectively be negotiated. “If you do not agree to our terms, do not use the service.”
6. In all but maybe the most important contracts, people are comfortable with contracts being “stored somewhere else by someone else.”
7. People are comfortable with almost any form of signature or acceptance.
8. “Trust” in some form is essential in the contracting process.
Many, if not all, of these assumptions and hypotheses have been in play in one form or another throughout history. We have seen the development of trust authorities like scribes and bureaucracies, recordkeeping, dispute systems, and other procedures as humans evolved into people who contract. We are likely to see some of those approaches be adapted to and developed for non-visual contracting.
Here are eleven potential approaches to consider.
1. Simplification. Requiring parties to listen to audio versions of 100-page contracts cannot work, even if AI is listening and interpreting for us. In non-visual formats, simplification is required almost by definition. Simple smart contracts are an example of a tool that might be used. Consider the example of a classic third-party escrow agreement. A smart contract could be triggered by the passage of the set number of days and send a payment to the seller. Smart contracts could be coded into some transactions without the need to see them.
2. Medium-specific Approaches. For example, voice recognition could authenticate a party to a contract on an audio device and be stored to evidence agreement. A few specific or important terms might also be confirmed by voice.
3. Reference. The governing terms might be displayed somewhere else through a hyperlink or provided by email or a portal either before or after the “signing.” This approach breaks down if a party has no access to visual media.
4. Standardization. The Open Source software licenses and the Creative Commons content licenses provide examples of using standardized and non-changeable terms.[4] The creator applies a license, and the user accepts the terms. The licenses become standard in the industry. Often, in law, lawyers work from standard templates from certain types of contracts. Those common templates might be standardized. One of my favorites examples is residential real estate contracts in St. Louis, Missouri, which were put together by realtors and the legal community many years ago. Home sales in St. Louis now rarely require a lawyer.
5. Limitations. Legislatures or regulatory bodies might limit what terms could be put in contracts and still be enforceable. We saw that approach with contracts of adhesion and racial exclusion clauses in subdivision agreements.
6. Implied Terms. The Uniform Commerce Code implies reasonable terms when specific terms are missing from contracts. A similar approach could be applied to non-visual contracts, to the extent the UCC might not already cover them.
7. Uniform Acts. As we saw with the Uniform Electronic Transactions Act and the Uniform Computer Information Transactions Act, the Uniform Acts approach used to create uniformity among the laws of the states might be a way to address the issues involving non-visual contracts.
8. Agents. The idea of intelligent agents on the Internet goes back a long way. What if we set up the terms we would and would not agree to and, when we made a purchase, our browser, app, or the software agent on its own interacting with the seller’s website automatically and invisibly negotiated and approved terms for us? Sound farfetched? It was discussed as a real possibility in the 1990s and is reconsidered occasionally.[5] (See, e.g., Project VRM, https://cyber.harvard.edu/research/projectvrm)
9. Platform agreements. I consider this approach to be highly likely as we move forward. When we join a platform (Amazon, World of Warcraft, etc.), we agree to a basic set of terms and the ways in which new terms can be added or changed on the platform. Everything after that happens seamlessly and invisibly.
10. Staying with the Status Quo. It is possible that lawyers will continue to force parties to enter into long, written agreements that require disputes to be resolved by courts. For a while, the status quo will remain part of the mix. Unfortunately, we have already seen difficulties raised by inconsistent court decisions and inadequate understanding of underlying technologies as we moved from paper to digital contracts.
11. Layering and Hybrid Approaches. I strongly doubt that we will see a one-size-fits-all solution for non-visual contracting. We will likely see combinations and layers of approaches to address the unique issues of each medium and the challenges posed in each. As we continue to experiment, that layering method will continue to change and evolve as the technology and how we use it continues to change and evolve, probably at an increasing pace.
Some Challenges We Face
A nearly universal belief today is that technology is always out ahead of the law and the legal system. That lag is accentuated by the number of years it now takes for courts to make final decisions on cases that are filed, the small possibility that legislatures will pass substantive laws in the current political environment, and the arduous process of getting regulations in place and enforced. The examples of law shutting down technologies are rare, with the Napster being the one most people will remember. Even in 2023, it is difficult to see law catching up with developments in artificial intelligence.
We will continue to see technology evolve at a rapid pace and stay ahead of law. Our current approaches will be challenged by a marketplace that is increasingly global and online. Especially with the Internet of Things, we can expect to see more non-visual user interfaces, many of which will be automated. As a point of reference, a quick Google search suggests there are between 14 and 43 billion devices attached to the Internet in 2023. That is an Internet of very many things. At each step, this technological change forces us to examine our core assumptions and principles about how we have always done things.
Let me leave you with a smart refrigerator example. Imagine that you buy a new refrigerator with all kinds of sensors, and it is connected to the Internet. It can sense when your milk is spoiled and order a new gallon of milk from your favorite local grocery store, pay for it, and have it delivered to you. And the refrigerator is totally controlled by voice.
How do you enter into a contract for this refrigerator and with whom? This smart refrigerator is an example of a “product as a platform,” a combination of goods and services using multiple providers that might or might not all interact with each other and could change over time. Would your sales contract with the refrigerator manufacturer cover it all? How would you contract with all the providers? How would that be done via a voice interface? This technology has been here for a while. Consider how this contract is handled today. Can we do better?
The Paths Forward
I use the word “paths” in the heading to this section intentionally. Addressing these challenges will require the legal profession and its clients to explore many approaches. Fortunately, while not easy, taking on this work requires many of the skills lawyers are especially good at: creatively understanding core issues and principles, building on what has worked in the past, understanding systems, and being willing to experiment to find the best results. There are lawyers already working on these issues.
The Internet tends to route around friction. Our current approaches, as applied to non-visual contracts, will cause friction in contracting and marketplaces. Will lawyers shape and participate how these transactions are done, or will the Internet be used to route around lawyers? How will they do that?
My best guess is that clients with push lawyers to face these issues. We will see experimentation begin to accelerate, using the scientific method and focusing on data and outcomes. I expect to see combinations and layering of the approaches I suggested in this article. Most important, we must increase the level of trust in transactions as technological change continues to accelerate. As Malcolm X said, “the future belongs to those who prepare for it today.”
Dennis Kennedy is the Director of Michigan State University’s Center for Law, Technology & Innovation. He retired as Senior Counsel, Digital Payments and Labs, at Mastercard. He wrote the book Successful Innovation Outcomes in Law, co-hosts The Kennedy-Mighell Report podcast on legal technology, and writes, speaks, and teaches on technology law and legal technology topics.
[1] See Electronic Transaction Act, Unif. Law Comm’n, https://www.uniformlaws.org/committees/community-home?communitykey=2c04b76c-2b7d-4399-977e-d5876ba7e034 (last visited Apr. 6, 2023).
[2] See generally Ill. Legal Aid Online, https://www.illinoislegalaid.org; see also Amanda N. Marino, Illinois Court Help: A Case Study in Court Customer Service (310), Legal Evolution (June 12, 2022), https://www.legalevolution.org/2022/06/illinois-court-help-a-case-study-in-court-customer-service-310/.
[3] See Linkedin, https://www.linkedin.com/in/tessa-manuello/ (last visited Feb. 12, 2023) (Tessa Manuello’s profile); Linkedin, https://www.linkedin.com/in/margaret-hagan-a7b18941/ (last visited Feb. 12, 2023) (Margaret Hagan’s profile).
[4] See generally Open Source Initiative, http://www.opensource.org (last visited Feb. 12, 2023); Creative Commons, http://www.creativecommons.org (last visited Feb. 12, 2023).
[5] See, e.g., ProjectVRM, Berkman Klein Center for Internet & Soc’y at Harv. U. (May 4, 2021), https://cyber.harvard.edu/research/projectvrm.
Any reproduction of the Article, including, but not limited to its publication, posting, or excerption in print, or on the internet, shall give attribution to the Article’s original publication on the online MSLR Forum, using the following method of citation:
“Originally published on Apr. 6, 2023 Mich. St. L. Rev.: MSLR Forum.”