Research and Advances
Computing Applications Contributed articles

Are We Free to Code the Law?

We should be, for the sake of millions of people with pressing legal needs.
Are We Free to Code the Law?, illustration
  1. Introduction
  2. Key Insights
  3. Assistance and Authorization
  4. Policy
  5. Freedom
  6. Legality Broken
  7. Acknowledgments
  8. References
  9. Author
  10. Footnotes
  11. Figures
Are We Free to Code the Law?, illustration

The emergence of interactive online services for legal self-helpers has triggered suppression efforts by the legal profession, as well as by state government officials in the U.S. While couched in terms of consumer protection, and at least partly motivated by such concerns, these efforts are also seen by some as blatant turf management by a profession anxious to avoid further erosion of its monopoly over legal advice and representation.

Back to Top

Key Insights

  • Online document-preparation services and other forms of automated legal assistance raise concerns about the unauthorized practice of law.
  • Such concerns should be balanced against social policy and economic freedom.
  • Software programs are more like books than like personal human services when determining whether they deserve protection under provisions like the First Amendment of the U.S. Constitution.

Often neglected in these discussions is whether restricting the distribution of software is within the legitimate scope of government action. No one would contend that attempts to suppress books, pamphlets, and speeches on how the legal system works and what forms one needs to interact with it would pass constitutional muster. Is providing software that helps people meet their legal needs an activity the state can prohibit under the U.S. Constitution?

Here, I explore ways software-based legal-assistance systems can be understood for purposes of public policy and constitutional analysis. The focus is on circumstances in the U.S., but many other countries face the same issues.

Back to Top

Assistance and Authorization

Individuals and organizations who need to prepare documents with legal significance turn to a variety of sources, including form books, courts, government agencies, physical form suppliers,a packaged software,b online form sites,c free online document repositories,d notaries public, legal-document technicians, conventional private law practices and corporate law departments, and virtual law practices.3

An increasingly popular, and controversial, category of service providers generates customer-specific documents over the Internet, using interactive software, without purporting to be engaged in the practice of law, including:

  • Commercial services;e
  • Nonprofit sites;f
  • Governmental and court sites (such as self-help court resources);g and
  • Free services by law firms.h

Most of these services use specialized document-assembly software long used by lawyers themselves; for an overview of document assembly and other specialized technologies used by lawyers, see Lauritsen.6 That technology enables someone to program “what words go where” under various sets of answers, gathered in interactive questionnaires that change as users work through them, with context-specific guidance. Applications can embody rule sets of arbitrary size and complexity and generate highly tailored and precisely styled documents.

In addition to commercial, governmental, and nonprofit initiatives, courses are offered at a growing number of law schools, some under an “Apps for Justice” rubric, in which students build useful software applications as part of their education, results of which can be made available to the public.i

The debate. Consider the following imagined example of the arguments one encounters (sometimes within a single head):

Voice A. At least in my state, these new services are blatantly illegal. By telling people what legal documents they need, and preparing them, they are engaged in the practice of law in all but name.

Voice B. Even if the provider makes perfectly clear it is not practicing law and the user explicitly acknowledges it?

A. We don’t think it is OK for unlicensed people to perform medical procedures, just so long as they do not claim to be doctors. Or to manufacture devices for self-help surgery.j

B. Cutting yourself open and generating a simple will are not exactly analogous.

A. Would you allow people to extract teeth and fill cavities without a license, so long as they do not claim to be dentists? What about self-help pharmacies that dispense drugs after some interaction with a medical expert system?

B. That’s different. Online legal help systems just provide information. Words. They do not do anything physically.

A. An improper legal “procedure” can cause a lot of financial and emotional damage, maybe even result in loss of shelter, child custody, citizenship, or liberty.

B. Lots of things people are allowed to do are dangerous. A weekend do-it-yourselfer can cause real damage with a power saw. Should we bar home-improvement television shows and limit power tools to licensed craftsmen?

A. Law is special. We need lawyers, and it’s only fair that in exchange for the years of education they are required to have, and the ethical rules they are required to follow, they get exclusive rights to perform certain kinds of services.

B. Come on. We have a huge population unable to afford legal help. Even unemployed lawyers are unwilling to work at rates low enough, and legal aid is grossly underfunded throughout the U.S.

A. That does not mean vulnerable people should be victimized by companies out to make a quick buck or even by well meaning do-gooders. Software rarely does justice to people’s legal needs.

B. Why should consumers incur the inconvenience and expense of hiring a lawyer to create documents that someone else is willing to do inexpensively or free? When they are informed of risks, and prepared to accept them? We are talking willing consumers here. This sounds like the nanny state.

A. We regulate many consumer transactions.

B. It seems to me that writing software is like writing a book, an expressive act that should be protected as speech.

A. Do not try to hide behind the First Amendment. These are not “publications” but services, with people behind them.

B. There are people behind books, too.

A. Yeah, but books don’t do anything.

B. Well, they do inform people, and they can be written to give very specific advice for very specific circumstances.

A. That doesn’t mean software deserves the same protection as written books.

B. Antipathy to these kinds of applications comes mostly from biased and techno-illiterate policymakers. Many lawyers, judges, legislators, and regulators have little understanding of the nature of computer code. And professionals naturally resist demystification of their expertise.

A. Stuff like this could destroy the legal profession. Is that what you want?

B. Hey, some of my best friends are lawyers. Lawyers just need to learn to compete on the merits. If machines can perform better than they can, they should consider another career path. Welcome to capitalism.

And so on…

The questions. The questions here fall into two groups: those about the power of government to regulate automated legal assistance and those about the wisdom of doing so. That is, can government prohibit automated legal assistance, and, if it can, to what extent should it?

Do people have a right to write, read, and run software that embodies ideas about how the law works? To what extent are people free to provide automated legal assistance? Is there a right to receive such assistance? To what extent can government enjoin or punish such provision or receipt? Is the distribution of software that helps people with their legal needs an activity that needs to be “authorized?” What is the right regulatory response? Is it good policy to forbid automated legal assistance? Should lawyers be given a monopoly over legal software, as well as over in-person legal services? In general, what are the appropriate boundaries? What principled lines can we draw in this area?

Unauthorized practice of law. Most states have defined law practice, as well as its unauthorized variants, in statutes and case law. Most such definitions extend to the selection and preparation of documents.

Attorneys General, bar authorities, and private plaintiffs in the U.S. have initiated proceedings against providers of automated legal assistance. Several matters are mentioned here to illustrate.

In the Parsonsk case, the Texas Unauthorized Practice of Law Committee sued two manufacturers of software that helped people prepare wills and other documents, and was granted summary judgment by the court. The case was mooted when the Texas legislature crafted the following statutory exception:

“In this chapter, the ‘practice of law’ does not include the design, creation, publication, distribution, display, or sale, including publication, distribution, display, or sale by means of an Internet website, of written materials, books, forms, computer software, or similar products if the products clearly and conspicuously state that the products are not a substitute for the advice of an attorney.”l

Is an occasional harm sufficient reason to forgo the power of modern information technology to make a dent in the vast unmet need for legal assistance?

In the Reynosom case, the court found a provider of software for bankruptcy preparation was engaged in UPL, laying stress on the point that websites are “put together by people.”

Many state bar committees have opined on this subject; for instance, in March 2010 the Pennsylvania Bar Association Unauthorized Practice of Law Committee concluded as follows:

“It is the opinion of the Pennsylvania Bar Association Unauthorized Practice of Law Committee that the offering or providing [in Pennsylvania] of legal document preparation services as described herein (beyond the supply of preprinted forms selected by the consumer, not the legal document preparation service), either online or at a site in Pennsylvania is the unauthorized practice of law and thus prohibited, unless such services are provided by a person who is duly licensed to practice law in Pennsylvania retained directly for the subject of the legal services.”n

That is, according to authorities in at least some states many of the services in the section on automated legal assistance are violating the law.

Back to Top


The case for prohibition. Arguments in favor of disallowing automated legal assistance generally involve protection of the public and of the legal profession:

Protecting the public. Some people will undoubtedly be harmed by automated systems. Defective or incomplete legal assistance can cause significant damage, and it is reasonable to assume such damage is more likely when no lawyer is involved.

Software applications lack common sense. They cannot hear what is not being said. They do not detect nuance or emotion. On the other hand, as with people, they can operate on unspoken assumptions, create the illusion of expertise, and engender unwarranted trust.

Protecting the legal profession. Lawyers are bound by many restrictions on their behavior in exchange for licensing. Is it unfair or unwise to restrict what non-lawyers can do in relation to giving legal advice, counseling, and representation? Part of the societal bargain regarding any profession involves a limited monopoly.

By not allowing unqualified people to advise citizens on their legal affairs, and seeing to it that such advice occurs within appropriately structured and protected relationships, we help ensure the smooth functioning of the legal system and the preservation of an independent legal profession that is so important to democracy.

The case for toleration. Those who favor allowing automated legal-assistance systems generally claim they yield net benefits for both society and the legal profession.

Given the vast amount of textual material already available to legal self-helpers, much of uncertain quality and with few clues as to currency and relevance to specific situations, interactive systems seem more likely to reduce harm than cause it. Their development requires significant time and money few organizations would invest recklessly.

Lawyers themselves are not infallible. Much legal work can be scripted, and software will eventually make fewer mistakes in many contexts. Machines have proven demonstrably better in certain law-related activities (such as coding documents for relevance to pending litigation).2

Counterbalanced against the inevitable harms automated assistance sometimes engenders are many clear benefits: more-informed citizens; better-prepared litigants; and cleaner and more-complete documents.

There are also considerations of economic freedom. Business and social entrepreneurs are anxious to innovate in the legal field. Threats of unauthorized practice claims chill innovation. An open market is the best defense against poor quality.

Reaching a balance. Do concerns about harms to consumers and the legal profession outweigh the benefits of citizens having access to legal knowledge through interactive programs? Is an occasional harm sufficient reason to forgo the power of modern information technology to make a dent in the vast unmet need for legal assistance?

The free flow of automated systems seems to offer net advantages. Reasonable regulations should be established to minimize potential harms, but a robust and open market of interactively coded legal ideas is in the best long-term interest of both society and the profession. It is desirable to have lots of such programs competing for use in a free market and to incentivize legal knowledge codification and systemization.

Imagine if a trade union of human “computers”o in the 1940s had successfully thwarted the development of electronic machines as the “unauthorized practice of computing.” We at least would not, I think, have to worry today about machines doing legal work.

Back to Top


Even if a good case could be made for regulating creation and distribution of automated legal-assistance systems, do such regulations pass muster under the First Amendment?

Admittedly, these applications are novel artifacts not envisioned by the founders.

First Amendment protections are not without exceptions; for instance, they do not authorize people to violate intellectual property or reputational rights. U.S. citizens are not free to engage in libel, copyright infringement, or sedition. Obscenity is only partially protected.

None of these exceptions apply to the expressive activity involved in automated legal-assistance systems.

Alleged misinformation or harmfulness is not viewed as justifying suppression of books, except in extreme circumstances. Government is not appropriately in the business of judging the quality or content of speech. A landmark casep held that distributing the 1965 book How To Avoid Probate did not constitute the unauthorized practice of law.

One may be inclined to suggest that some automated systems are a form of “commercial speech” and thus deserve less protection. Commercial speech has generally been understood as the activity of beckoning business, not the substantive content of what is being offered. Selling a book does not render it any less deserving of First Amendment protection than giving it away for free.

An alternative way to avoid First Amendment issues is to conclude that programs are not “speech” at all but a form of conduct, analogous to the work of manual document preparers. This involves distinguishing between “pure” speech and “speech plus” that entails actions, as well as words. Sometimes speech-related action is not protected if it is physically dangerous. Does such a dangerousness rationale extend to communicative action?

Several legal scholars have tentatively concluded for the unconstitutionality of repressing online legal services under the guise of the unauthorized practice of law.q The following sections lay out an analytical framework that may support more definitive conclusions.

A typology of expressions. Figure 1 outlines one way to organize the varieties of expression a legal self-helper might access; blue boxes are categories, and green boxes contain examples.

Expressive conduct falls into two main categories: creating artifacts, or works of authorship, and “performing,” or engaging in live, real-time communication with others. Artifacts in turn are either static (with fixed content in a fixed order) or dynamic (programmed to present different content in different orders depending on external triggers (such as a user’s behavior interacting with it). Performances fall into two high-level categories: those in which communication is unidirectional, or one-way, (such as speeches) and those in which communication is bidirectional (such as one-on-one and many-to-many conversations).

Some features apply to multiple branches of the Figure 1 tree:

  • Most modes of expression can be through either physical or electronic means; for practical purposes, programmed content and social-media interaction can be accomplished only electronically;
  • Electronically mediated expression can happen offline or online; that is, via electronic networks (such as the Internet) and protocols (such as the Web);
  • Artifacts can include charts, diagrams, tables, flowcharts, decision trees, and other graphical elements; such things can also be used in most forms of performative expression;
  • Artifacts can include audio and video elements that can also be used in performances; and
  • Artifacts can include structural and navigational features (such as tables of contents, indices, and links); with physical artifacts the reader does the work; in electronic ones, buttons and hyperlinks make navigation easier. Many artifacts involve arbitrary access to any part (such as by page turning, fast-forwarding, and scene selection).

Are software programs more like books or like human services? The difficulty of reaching a satisfactory conclusion about automated legal assistance arises in part from our instinctive assent to two propositions:

  • People should not be allowed to do through a program what they are not allowed to do in person; and
  • People should not be disallowed to do through a program what they are allowed to do through books and other media.

To the extent a software application is viewed as a kind of personal conduct, it makes sense to apply the treatment one would apply to comparable functions being accomplished through an in-person service. To the extent a software application is viewed as a work of authorship, it makes sense to apply the treatment one would apply to the comparable content delivered through a book. How can these competing views be resolved?

We might first acknowledge that software applications are a tertium quid, or something similar to but distinct from both books and services. Like the wave/particle duality of light in modern physics, perhaps it makes sense to regard software as both a “work” and a service; Figure 2 outlines the shared and unshared characteristics of these kinds of things.

Software programs share characteristics with both books and instances of service delivery. Like books, they are essentially textual works of authorship, fully written in advance of their use; the author is not present at the time of use. Like services, they can be dynamic, bidirectional, and generative (such as by producing case-specific answers and documents). Unlike both, programs operate as machines, with automated behavior, and are rule-governed and deterministic.

Any of these modes of communication can be used for the transmission of knowledge, guidance, opinions, and expertise. The content being delivered can be “neutral” or tilted in favor of a particular kind of party or point of view.

Programs as texts. When in use, software applications typically involve no contemporaneous human involvement by their authors. Users interact with pre-written code, with no other human interacting with them as they do so.

Programs are special forms of words and numbers, textual objects that instruct machines how to behave. Any program can by definition be expressed textually. You can think of them, as hypertext pioneer Ted Nelson put it, as “literary machines.”8

All outputs of an automated legal-assistance system are also in the form of textual speech acts. Delivering a document someone can download is not meaningfully different, except in terms of convenience, from presenting content that in effect says, “Here are the words you need, in this order.”

That is, these systems not only emit texts, they are texts.

While debate among legal scholars continues as to whether the First Amendment extends to “symbolic” speech like flag burning,10 there is little doubt it protects written texts. If I have the right to share the text of a program with others, and they would commit no offense by compiling and running it, why should I not have the right to run the program and give them access to it?

The question of whether First Amendment rights extend to computer code has arisen in cases involving publication of decryption algorithms; for example, “[C]omputer source code, though unintelligible to many, is the preferred method of communication among computer programmers. Because computer source code is an expressive means for the exchange of information and ideas about computer programming, we hold that it is protected by the First Amendment.”r

Back to Top

Legality Broken

Like the world that inspired gamers in Jane McGonigal’s 2011 book Reality is Broken,7 the legal system in many countries is broken in many respects. Millions of people with pressing legal needs go without help. Courts are underfunded and overwhelmed. Many lawyers are unemployed or underemployed. Some law schools are struggling to survive. Recent law graduates are drowning in student loans.

Forbidding distribution of self-help legal software is not only of dubious wisdom as social policy, it is offensive to First Amendment values. It is difficult to make a principled case for suppressing freedom of expression about how the law works.

Free expression by definition need not be “authorized.” Honest attempts to transmit knowledge about how the law works should not be suppressed, at least when done in ways that do not impersonate trusted lawyer/client relationships. Free citizens should not be required to have a license in order to express their understanding of how the law works or to sell or give away such expressions.

Coded law is not something, like hate speech at a military funeral, we should have to tolerate due to concern for higher values. It is an affirmative good we should embrace.

It is in the enlightened interest of lawyers, as well as the best interest of society in general, to enable programmatic expression of legal knowledge. We should be free to write code, run code, and let others run our code. If concerned citizens, law students, and entrepreneurs want to create tools that help people access and interact with the legal system, the government should not get in the way.

Are citizens at liberty to create and share software that helps others understand and interact with the legal system? Are we free to code the law?

We certainly should be.

Back to Top


Research for this article was partially underwritten by I am also grateful to Pamela Samuelson and Laurence Tribe for their helpful suggestions and reactions.

Back to Top

Back to Top

Back to Top

Back to Top


F1 Figure 1. A typology of expressions.

F2 Figure 2. Three modes of assistance.

Back to top

Join the Discussion (0)

Become a Member or Sign In to Post a Comment

The Latest from CACM

Shape the Future of Computing

ACM encourages its members to take a direct hand in shaping the future of the association. There are more ways than ever to get involved.

Get Involved

Communications of the ACM (CACM) is now a fully Open Access publication.

By opening CACM to the world, we hope to increase engagement among the broader computer science community and encourage non-members to discover the rich resources ACM has to offer.

Learn More