Roads, Towers, and Online Legal Help

Online legal help systems have been a major force for good. Millions of people without the luxury of a personal lawyer have benefited from them. They come in many shapes and sizes, from many different worlds:

· Legal aid programs and other nonprofit legal service organizations

· Courts and government agencies

· Law schools and universities

· Commercial providers and startups

· Private law firms and departments

We have an embarrassment of riches when it comes to tools, platforms, and methodologies for applications that address legal needs. The diversity is a sign of health. But there’s also a lot of suboptimal duplication of effort and missed opportunities around scale and synergy. The map of coverage also remains very sparse. For most people, in most situations, there’s no immediately useful online resource, free or paid.

Where is this all going? Where should it be going?

(My focus here is on not-for-profit efforts, in North America. Commercial and international developments of course make this even more interesting.)

A thought experiment

Those who venture to supply online legal applications share many challenges.

Getting content correct and keeping it current with limited human and financial resources is a big one. Applications need care and feeding. Users and developers need support. But there’s a shortage of organizational bandwidth.

Imagine if we organized a shared resource to ‘mutalize’ some of the infrastructure needed by multiple content providers. What might such a resource look like?

It would supply secure and scalable servers that application providers could use, without having to source, configure, and manage their own.

It would provide an organized and accessible content collection, optimized for distribution and sharing, and cover more than one development platform.

It would offer accounts for end users, developers, and content managers, with associated data storage and sharing (across time, apps, and people). Users’ answers to questions would be securely stored as tagged sets of data that can be accessed and edited by multiple applications.

It would support users via email and live chat, bug tracking, and ‘ticket’ management.

Such a service would provide training, continuing education, webinars, and navigable knowledge bases of relevant materials for developers and other project participants.

It would help build community by hosting discussion fora, regular online meetings, and other arrangements that facilitate collaboration.

It would offer one-on-one technical support to developers and managerial support to project teams. Custom statistical reports about usage would be available via a self-help dashboard.

This imagined resource would support multiple human languages, integrate with external systems such as case management and e-filing, and offer specialized configurations for in-person and virtual clinics at which groups can be served simultaneously or asynchronously.

You might think of it as an ecumenical collection of free legal apps with answer-saving and other valuable forms of automated assistance.

LawHelp Interactive

It turns out that we already have at least one example of a service that exemplifies all of the above qualities — LawHelp Interactive (LHI).

LHI was first envisioned over 18 years ago. Planning began in late 2001. It arose in part from the ashes of AmeriCounsel, the dot-com adventure in which my colleague Bart Earle and I first gained experience with a large scale online document assembly deployment. LHI started out as National Public ADO (Automated Documents Online. Get it?) Pro Bono Net, a national nonprofit dedicated to access to justice, assumed responsibility for the service in 2006 and soon came up with a better name.

(For more about the ambitious AmeriCounsel venture, including its Open Practice Tools initiative, check out my keynote at the 2001 CALI conference.)

The federal Legal Services Corporation has played a central role in supporting LHI. It provided seed funding for initial R&D, and later acted as a strategic investor, supporting ongoing operations and improvements, as innovations in service delivery to unrepresented litigants and others were proven out. This has been a successful public/private model; today LSC provides only about half of LHI’s budget.

LHI has been foremost — at least in terms of scale and impact — among free US resources that leverage the web and intelligent technology to advance access to justice and legal wellness. It has accumulated some 5000 modules, is used in over forty states, and offers its interface in seven languages. Millions of sessions have resulted in millions of customized documents. All without charge. (LHI is free for end users and nonprofit legal aid programs funded by LSC. Courts can subscribe for access.)

LHI has been a real catalyst for innovation. The program and community have pioneered new models of access to justice centered on online forms in many states that have dramatically improved the ability for those without lawyers to achieve justice on their own, including unbundled, limited scope services and remote services. (See https://www.nytimes.com/2016/06/01/opinion/legal-aid-with-a-digital-twist.html and https://www.srln.org/node/847/srln-brief-examples-lhi-driven-partnerships-probononet-2016.)

A substantial percentage of LHI usage is by lawyers, paralegals, advocates, court staff, and other professionals.

From the beginning LHI was conceived as including a content commons, a collection of codified legal know-how that could be freely copied and remixed by participating contributors. A partnership with the Center for Computer-Aided Instruction (CALI) cemented an early determination to support multiple interfaces for end users. A2J ‘guided interviews’ have been part of LHI since the get go.

Driven by a spirit of continuous improvement, many enhancements have been delivered. There’s been steadily increasing geographic and functional scope, including custom integrations with legal aid and court software systems. (See e.g. https://www.connectingjusticecommunities.com/ny-courts-honored-with-ltn-innovation-award-for-the-most-innovative-use-of-technology/2015/08/.)

More advanced forms of process guidance, via dynamic ‘pathfinders’ and ‘next steps,’ are in the works. A recent survey identified over thirty further potential enhancements for community prioritization.

So LHI is going strong and still growing. Yet it faces challenges. Its UI is periodically refreshed, but aspects feel old-fashioned. Part of that is due to needing to attend to an installed base, and gingerly manage legacy code. Complex services tend to accumulate technical debt. That makes re-architecting things a tall order. (For example, presently all files for each LHI module live in their own container, making it difficult to do quick global updates of textual or logical components used by more than one.)

Another concern is LHI’s primary document assembly engine.

Two months after the 9/11 attacks a white paper on online document automation commissioned by the Legal Services Corporation was circulated. It described the already mature and variegated market of products and their potential use for legal assistance purposes. Twelve technology providers were then asked to complete a detailed questionnaire about their solutions. Six responded.

The next February (2002), several dozen folks from around the country gathered in a conference room at Davis Polk in New York City to hear extended presentations from four vendors. Alternatives were reviewed and assessed. HotDocs emerged as a solid candidate and was adopted by the initial team. It has proved reliable for a couple decades. It still has a huge domestic and international customer base. But there are downsides.

HotDocs has undergone several changes in ownership since it was generously donated for use on LHI by LexisNexis. The latest acquisition, by AbacusNext, has reminded us of the vicissitudes of depending on a commercial vendor. Pricing has become more aggressive, although the company has confirmed a commitment to a 70% discount for nonprofits. And most of its development energy has been allocated to next-generation products that aren’t yet a good fit for LHI’s content and community, which depend on its ‘classic’ line.

The HotDocs authoring tools are not free, and require Windows. That’s a particular constraint for law students, many of whom use devices only running the Mac operating system. (You can run HotDocs interviews on just about any device, but need Windows to create and edit them.)

Also, while the JavaScript interface offered by HotDocs is highly functional and stable, it looks increasingly outdated.

LawHelp Interactive may seem like an incumbent surrounded by disruptors, but the future needs something like it. Like a caterpillar, LHI has morphed several times from early conceptions. What butterfly might emerge next? Where does it go from here?

(The rest of this article lays out a bigger context in which LHI will likely play a role. I don’t purport to speak for the project or Pro Bono Net.)

The spiraling ecosystem

In the meantime, others have brought new ideas and energies to this space.

Odyssey Guide & File from Tyler Technologies emerged as a commercial alternative to CALI’s A2J, aimed at courts that want to field interviews and electronic forms to simplify the filing process for self-represented litigants.

For its part, CALI has implemented native document automation features (it previously relied on HotDocs to assemble documents), and moved to independently hosting guided interviews on its own A2J.org site.

An excellent open source alternative finally arrived, in the form of Jonathan Pyle’s docassemble. (See Making Mischief With Open-Source Legal Tech for an example of its power.) And several nimble players jumped in to make available easier interfaces for building and maintaining docassemble applications, Community Lawyer and Documate. The former now boasts it own content collection — https://community.lawyer/search.

Legal document automation tools have long been a commodity. I’ve used dozens of them. Deep, reliable functionality combined with provider longevity is still rare. Free and easy go a long way, but maybe not far enough. HotDocs, for instance, includes so-far-unmatched facilities for automating complex sets of PDF forms. And history is littered with brilliant alternatives that failed to survive.

Spaces adjacent to document services have likewise been busy. The Civil Resolution Tribunal’s Solution Explorer, a pioneering expert system offering free legal information and tools, has been used over 100,000 times. New players are emerging regularly — see e.g. https://www.q4justice.com/.

Law schools have jumped into the game, with courses in which students build applications as part of their course work using tools like A2J Author, Community Lawyer, HotDocs, Neota Logic, and QnA Markup. (I’ve taught such courses at five different schools myself, and fellow teachers are active around the world.)

Commercial players and startups have also been making waves. Neota is seeing worthy competition from entrants like Bryter; Legal Zoom may be noticing ascent by disrupters like DoNotPay. There’s a bit of a space race going on. Major investments are happening.

Other platforms offer features and functions that LHI does not, but few have comparably robust fabrics of surrounding support. We’ve ended up in multiple camps, with mutually inconsistent tools and skills that are not readily transferred. This fragmentation is discussed more below.

New frontiers

There’s no lack of new things that LHI and related services can and should do.

One perpetual desire is to ease the authoring and maintenance of interactive content, especially by non-programmers such as domain experts. Some products and platforms have made authoring much simpler, at least for basic applications. This piece illustrates how vendors can trumpet ease-of-development advantages. Low-code and no-code are more than buzz words.

Many of these services fall short in terms of accessibility. Providing genuinely usable applications for those with serious cognitive or perceptual limitations via browser-agnostic Web sessions that need to present, elicit, and generate complex texts — even on smart phones — is an enormous challenge.

All services could expand their interoperability with each other and with common third party tools like Clio, DocuSign, Google Drive, and Legal Server. (LHI was one of the earliest to work with the latter. Other services have also done impressive work in this area.)

We should remain alert to new tools and paradigms for dynamic questioning, fact-specific guidance, and document generation. Those will include new forms of interaction — bots, text messages, and other conversational approaches. ‘Push’ — proactive communication of warnings and reminders to users — will also have a role.

And we should be on the lookout for tools that support new kinds of assistance, ones that go beyond interactive questionnaires, custom instructions, and assembled documents.

One of my candidates for a new field of endeavor has long been decision support. See A Decision Space for Legal Services and The Centrality of Choice in Legal Work. Tools that promote effective choices naturally also have usefulness in the online dispute resolution (ODR) context.

We can clearly find ways to introduce more artificial intelligence into our online legal help environments, both with ‘good old fashioned AI’ like expert systems and next generation deep learning and pattern recognition systems. We could intelligently parse documents both to help users in specific situations and to infer models for use more generally.

Open source and other forms of openness of course present great opportunities. See Opening Legal Knowledge Automation. Among other things, that could involve standardization of data elements and structures, shared ontologies, and variable namespaces.

One aspect of all this that I’ve been particularly vocal about is quality. We could use more fanatical attention to it, maybe in a six sigma, zero-defects spirit. See The High Cost of Quality in Legal Apps and Substantive Legal Software Quality: A Gathering Storm?. Quality assurance — regarding both platforms and their content — would be facilitated by greater transparency and inspectability.

The above is a very incomplete list. Which is both exhausting and encouraging. This piece from 2007 describes frontiers in legal document automation more generally, some of which remain uncrossed: Current Frontiers in Legal Drafting Systems.

Strategic choices

All of the providers face the challenge of funding and sustaining their efforts. And it’s hard enough just to keep things going; trying to rebuild while operating at capacity can be like changing engines on a plane while in flight.

One other shared concern is the specter of intensified regulatory scrutiny, e.g. by bar groups contending that assistance via software is tantamount to the unauthorized practice of law. My own view, articulated in places like Safe Harbors and Blue Oceans, is that there should be a bright line rule that making software available is not ‘practice.’ Which is not to say that we shouldn’t be concerned about the quality of some of the published content. However, we can deal with bad actors without resorting to prior restraint. Even if such constraint were constitutional it would be bad policy. That a work of authorship can be made to do useful cognitive work doesn’t deprive it of 1st Amendment protection. And freedom of expression is an empty promise without freedom to distribute what is expressed.

One key question is how we might best work together? Should we try to be less disintegrated than we presently are? What’s the right balance of centrality and distribution? What kinds of things are best accomplished together, and which happen best on the edge? Can we reap the benefits of cross-organizational thinking without losing those of autonomy?

Do we need a mother ship, or is a loosely coupled federation better? Are big shared environments good things? Is there a place for a neutral “Switzerland” of free interactive content, one that supports many application categories/types? Should content itself be platform independent?

Clearly we don’t want to put all of our eggs in one basket, and no organization should spread itself too thin or try to be all things to all people. Organizations of course should focus on their key commitments and core competencies. They also need to maintain continuity of teams.

Towering

As things now stand, centrifugal forces are in play, and lots of folks are off doing their own things, using different languages and approaches in multiple ‘towers.’ (We would resemble the tower of Babel if there was only one! I guess we’re more like a medieval Italian city-state, with wealthy families competing to have the tallest structure.) Lots of hard-won wisdom remains trapped in silos.

Some of this comes from ‘not invented here’ attitudes; some from welcome entrepreneurial zeal. But competition is generally healthy, even at the expense of wasteful duplication and reinvention.

And there are clear benefits of diversity in our ecosystems. (See Knowledge Gardening and Civil Justice Engineering.) It’s going to take a lot of villages (and villagers) to ameliorate the access to justice crisis.

But shared resources can offer economies of scale. Some of those economies can be tapped through loosely coupled arrangements. There would seem to be much positive opportunity in better connective tissue. That of course raises governance challenges. And any arrangement will naturally involve costs and compromises.

A facilitator of shared resources, collaborations, integrations, and interoperations could function as a benevolent natural monopoly. It could supply participants with valuable insight into each other’s collections, and maybe even interchangeable parts.

In the widening gyre perhaps such a center cannot hold, but let’s hope we won’t need to settle for mere anarchy.

What topology would make the most sense? No center, one center, or multiple centers? If center(s), what form could such entities take? Or is this all best left to the market?

These would be good topics for a hackathon, or at least a designathon. Perhaps this socio-technical challenge will come up at this year’s SubTech conference.

On the Road

We tend to take streets and highways for granted. They’re actually quite impressive structures, albeit not very high. Ribbons of pavement wind through our neighborhoods, cities, and countrysides. They provide the immensely valuable commodity of safe, level ground on which arbitrary kinds of vehicles can move. Heterogeneous traffic flows on the interstate highway system with relatively little friction.

Road building and maintenance don’t offer much glory, but their effective practice is critical to modern life. Commerce needs a commodious and reliable substrate.

We’re blessed with lots of bright ideas and cool tools in the access-to-justice world. Inventors and visionaries remain essential. But we also need planners and managers. Institutions can help. To revisit an old trope, we may want a cathedral as well as a bazaar. Nimble authoring systems are great, but we also need distribution systems.

Engineering a Better Tomorrow

Last spring researchers at Cambridge University announced that they had synthesized the complete genetic material of the bacterium Escherichia coli — four million base pairs of DNA — and inserted it into functioning cells, which reproduced and survived. Quite a feat of engineering. Could legal technologists accomplish something as impressive?

We’ve long had the tools and knowledge to help under-resourced people deal effectively with their legal needs. But we haven’t yet made a major dent in the problem. We could do SO much more. The need for these kinds of services is easily 100 times greater than what is presently provided. We’ve only scratched the surface of positive potential.

Unsolved legal problems cause immense suffering. With adequate help, that suffering can often be avoided, or at least minimized. Yet many folks get little or no help, even to help themselves. Many go unrepresented in formal proceedings; even more are totally ‘unhelped’ across the vast range of law-related problems and opportunities.

Vendors come and go; projects rise and fall. But aching needs remain.

How about an Apollo program to end poverty of legal help? Imagine we were taking the best emerging technologies and accumulated know-how to address this need. How might we sculpt and nurture a system of systems that stands a chance of achieving that result?

Our goal should be nothing less than the eradication of legal helplessness. A vibrant market of high-quality, reasonably priced services, supplemented by equally high-quality resources for those who can’t afford to pay. Those who need and want help should be able to get help.

That moonshot will require decent roads as well as sturdy towers.