What real transformative change could look like

Computational law. Documents have become computational instruments. Whether a contract, a plan, policy, memo, or other instrument, documents are written so that computers can execute them. This doesn’t mean that all lawyers must become programmers. Lawyers will work with experts in coding documents. It does mean that documents serve two purposes. First, as computational instruments, they remove from humans the need to interpret every provision of the document. Whenever possible, the document is “self-executing” removing humans from the loop. The computer can “read” and execute provisions within the document. Second, human-readable versions of the document are available (assembled by the computer) on screen for those times when humans need to be in the loop. Computational does not imply entirely removing humans; it does mean removing humans when they are not necessary.

As computational documents, payment, terms, and default triggers are all built into the document in a way that a computer can “read” and execute the provisions. Documents use XML (eXtensible Markup Language) coding as the standard for tagging language within the document. A computer can execute the code and generate an exception report if needed. All documents are written in the same standard code (XML, in my version of the future), so documents can be read and executed regardless of the computer platform used by a party or third-party. This interoperability is important. Legal materials can’t become isolated islands.

The same coding and interoperability concepts extend to laws. They are written (or re-written) using standard formats and XML tagging. This way, a computer can assess performance under a statute by comparing data to the statute’s requirements. (IARPA has funded seedling projects attempting to accomplish this with existing statutes.) Documents can incorporate a statute’s requirements by reference (subroutine) allowing documents to conform to legal requirements without having to re-write those requirements into the document.This one change saves tens of thousands of hours of lawyer time spent “translating” statutory language into documents. It also saves on compliance costs. In-house compliance teams can take compliance statutes, plug them into the company’s computer, and electronically compare behavior against the statute. Again, humans are still in the loop when needed, but outside the loop when not needed.

Using XML tagging allows parties to compare terms across many documents. Documents have three parts: public common, public unique, and private. The private part contains the confidential terms of the document (e.g., price). The public parts contain the non-confidential terms, broken into terms that are common among documents and terms that are unique to the document. Publishing the non-confidential part gives the user access to the non-confidential part of other documents, so users can compare terms across documents.

In addition to XML tagging statutes, court decisions are XML tagged. This allows lawyers to incorporate the language of a court decision by referencing specific language in a case. It also allows comparison across cases and jurisdictions. The law is no longer hidden or trapped in documents, it is available to all for analysis and critique.

Of course, all cases and statutes receive a unique identifier (much like today’s digital object identifier or doi). This unique identifier means each case and statute has a permanent ID not tied to any publisher. All legal materials from the government are in the public domain and published on digital ledgers (to ensure that only the authentic version issued by the court or legislature is identified by the unique ID). The law, for once, is readily available to all at no cost.

The combination of using standard coding (XML) and using it across all forms (statutes, cases, documents) interconnects the law. Parties who want parts of their documents governed by particular statutes, regulations, codes, or cases can do so by using the unique ID for the source in their document (e.g., parties can refer to the unique ID for the original statute or add language incorporating updated versions of the statute as published on the applicable digital ledger).

AI becomes a thing. Artificial intelligence becomes a tool to use with the digitally interconnected universe. AI can search through the thousands of documents, statutes, regulations and other materials to find patterns, similarities and differences. It becomes a sophisticated tool for searching through the growing mound of digital documentation. Much like doctors using AI to find new ways of looking at the incredible volume of data at their disposal, lawyers could finally do the same with legal materials.

Asking fundamental questions. Putting all legal information in digital form, annotated with XML tags, is a good first step in the radical transformation. But, it is just that — a first step. Digging further, we need to explore fundamental questions about our legal system. Remember that what we do today is based on 18th and 19th century concepts of how to address many problems. For example, our ideas about property date back to a time when property ownership and rights transfers were much simpler than today.

Today, we struggle with notions of who owns what. Real estate is relatively simple. But, intellectual property in an age when digital transmission and copies are simple presents some difficult problems. AI will complicate things. Can AI own property? If not, who is responsible when AI creates an infringing work? The AI owner? The programmer? Someone else? We all have strong opinions on these issues, but we don’t have clearcut answers. A radically transformed legal industry must address these issues.

What about simple torts. With modern vehicles producing gigabytes of data per second, proving liability in a simple tort case will become an exercise in big data. A radically transformed legal industry must address this new challenge to basic tort practice. Is there a presumption that the computer is correct? Do we even use traditional tort law, or do we use agency law as some have suggested? The court system must be updated to handle a digital world. For example, many disputes can be handled by computers using algorithms to resolve basic issues (defaults and remedies could be given various weights).

As you can see, radical transformation isn’t just about tweaking how legal services are delivered, it is about re-thinking legal services at a fundamental level. Many will find this unsettling. The law, they believe, is well-settled in many areas and should not be disturbed. But, if that law does not fit today’s reality, why should we tie ourselves to it? From a client’s perspective, why rely on laws that are out of date and can’t address modern legal problems, just because they represent “settled legal precedent.” Would you expect your doctor to stick with early 20th century notions of medicine just because they were well-settled? Would you want your doctor to seek out data on modern treatments? Why should the rules that govern the relationships between and among people be any different?

IoT and embedded law. We can run through many other changes that would represent radical transformation. The Internet of Things brings us devices with technology embedded in them. They also can have legal information embedded in them. They could form contracts (e.g., for repair services), signal potential problems, or perform different functions depending on whether the IoT owner is in good standing or default.

It’s all in the data. Data will become the new bedrock on which legal decisions are made. By annotating statutes, cases and documents, we enable AI to mine this vast treasure trove of data. How often is a statute relied upon? What are the outcomes of cases analyzing the statute? What are the common terms in manufacturing agreements? No more arguing about what is “market.” A simple search of public documentation will reveal how often a certain provision is or is not used.

How often to plaintiffs prevail in certain cases? What are the remedies typically awarded? Today, these are tough questions to answer requiring labor-intensive efforts. Tomorrow, an AI search will get the answer in minutes. Instead of the law being a mystery, we all will be able to see not only how the law is applied, but also the practical outcome of those applications.

Questions today that are routinely incorrectly answered (what are the chances of my winning in court) will become easy to answer with hard data. AI can slice and dice the publicly available legal data giving transparency to what previously was hidden. Both parties to a dispute will know the odds and risks before they get to the courthouse door. This alone will encourage parties to resolve their disputes rather than waste valuable time and resources litigating.

Processes will prevail. The focus on data and computational documents will heighten the focus on processes. This will enable law companies to take a significant lead over law firms when it comes to anything but truly bespoke matters. Law companies typically start from a strong process perspective and build that into their service offerings. Law firms have to retrofit processes on to what they do and face the enormous challenge of retraining lawyers. That is why I give law companies the advantage. Just as third-party IT providers and HR companies took over much of the day-to-day work of their respective in-house counterparts, law companies will take over the day-to-day legal work of law departments. The lawyers who remain in-house will work closely with the business teams to identify the needed or preferred legal services to support the business strategy, and then will turn over the routine work to support the strategy to the law companies. Law companies, providing similar services to many companies, will be far more efficient and cost-effective than in-house teams (e.g., law companies will invest in specialized software and the necessary support services).

Quality will significantly improve. The focus on streamlining documents and making them computational, processes, and simplification of law, will combine to improve the quality of legal services. Variability will be reduced, decreasing the chances of errors. Various safeguards can be built into the processes (e.g., computerized integrity checks of documents). Lawyer creativity will no longer be focused on making every document different. It will be focused on using data and other information to identify risks and develop appropriate mitigation strategies. In some cases, it will be focused on developing laws so that new industries can grow (e.g., drones).

Don’t forget individuals. While most of what I have described applies to corporations, life will be better for individuals. As documents become computational and streamlined, the focus of practitioners who serve individuals will shift. Instead of making money off document preparation, they will make money 1) through personalized counseling, and 2) through volume. This is where value pricing comes into play. Preparing documents will take only a few minutes, at best. But, counseling will require time. Set fees for certain services will put the emphasis on efficiency in routine work. Instead of handing me a stack of thick documents that encompass my will, trust, and powers of attorney, the documents will be digital and computational. A court, bank, or medical office will accept the digital document and be able to interpret its provisions (all digitally signed, of course). The overall reduction in cost per individual will open the market to more people receiving basic legal services. The lawyer may make less per person, but will be able to serve a much greater number due to lower costs. Differentiation will come through areas computers can’t handle, such as counseling and knowledge of the local market.

One view, not the view

Obviously, there are many aspects of legal services that I have not covered. But, you get the general idea of one direction that could result in radical transformation of legal services. This direction could result in far fewer lawyers, though some of the reduction may be offset by lawyers focusing on new practice areas (e.g., space law, drone law, IoT law). The key is that lawyers will give up doing things that add little or no value (endless tweaking of documents to meet personal quirks) in favor of true value-added services. Computers will take over doing the routine things that today absorb labor. While strange at first, the shift will feel natural in short order. It will free up time for lawyers to practice “at the top of their license” rather than perform rote work.

Equally obvious, this is one version of the future. I’m not arguing it is the only version or even the most likely version. I offer it as an example of what we could do with existing technology. Nothing I have described requires new developments or innovation. It requires changes in behavior and a willingness to focus on what benefits the client rather than the lawyer. It also would require changing how we train lawyers and the addition of individuals with new skills (e,g, those who could help turn documents, laws, and cases into computational instruments). Law departments could begin migrating to this new world, and law companies could assist them. A progressive law department could substantially reduce its costs within one to two years. But to do so, would require change.

And that brings us back to the fundamental problem facing the industry: those in the position to drive change see no rewards in doing so.