Regulatory Innovation With a Legal Tech Sandbox

On April 22, 2021, the Law Society of Ontario approved a “Regulatory Sandbox for Innovative Technological Legal Services”, a five-year pilot project through which non-licensee providers will be given the LSO’s blessing to provide “innovative technological legal services” directly to consumers, under the LSO’s supervision. The sandbox was recommended by the LSO’s Technology Task Force in its report released on April 13, 2021. The sandbox is currently slated to launch in October 2021.

The proposed regulatory sandbox emerges after over three years of study by the Technology Task Force, which the LSO established in February 2018. Although it has taken considerable time to get to this point, the proposal is a welcome development that will permit new, innovative, non-lawyer providers to offer legal services to the public, and which should provide useful insight for the future of legal services regulation.

This column will discuss what the sandbox looks like, why it’s a promising step, a few concerns I hope law societies will consider going forward, and what the future may hold.

I. What is a “Regulatory Sandbox for Innovative Technological Legal Services”?

As it currently stands, licenced lawyers (and licensed paralegals, in Ontario) are the only persons permitted to provide legal services to the public (subject to certain limited exceptions, discussed in Part III of one of my earlier columns).

There is also a serious and continuing problem with low- and middle-income access to justice in Canada.

Expanding the market by permitting new legal services providers should help fill the access to justice gap. But bad legal advice is likely worse than no legal advice—and the risks to the public of poor-quality legal services should be addressed.

A regulatory sandbox provides an opportunity to explore the provision of legal services by new providers to serve unmet legal needs, while putting safeguards in place to protect the public from being harmed.

The LSO’s proposed sandbox will exist for five years. Once it is up and running, the LSO will consider applications (including from for-profit and non-for-profit corporations) and, where appropriate, approve non-licensee entities to provide technologically innovative legal services directly to the public. Approval decisions will consider viability, benefits to consumers, and quality assurance mechanisms, among other things.

It is expected that each entity will operate in the sandbox for about two years, during which time they will be required to comply with “risk-based monitoring and reporting requirements”. While these have yet to be specified in detail (and in all likelihood will evolve during the course of the sandbox), they will focus on protecting the public from four key risks of harm:

  • Failure to exercise legal rights or pursue legal recourse as a result of ignorance, error, or poor-quality legal services;
  • Purchase of unnecessary or inappropriate legal services;
  • Exposure or sale of confidential client data to third parties; and
  • Inability to seek redress or recompense from a legal service provider that has failed to provide the service expected or agreed.

Participants will also be required to enter into a participation agreement with the LSO, which will include protocols for data collection and data governance, a mechanism to address user complaints, insurance requirements, and reporting and auditing requirements.

After a couple years in the sandbox, the LSO will determine whether, and under what conditions, participants may be permitted to continue providing legal services. This could be by providing individual permits to individual participants who satisfied performance objectives (presumably by demonstrating that they are serving consumer needs without creating undue risk of harm), or by adjusting regulatory requirements more broadly.

Ontario is not the first North American jurisdiction to do this. That distinction goes to Utah, which started implementing a regulatory sandbox for non-traditional legal services providers in August 2020, after a joint task force formed by the Utah Supreme Court and State Bar recommended the initiative in August 2019. Utah’s sandbox, overseen by a new Office of Legal Services Innovation (a division of the Utah Supreme Court), is well underway; as of April 1, 2021, it had approved 22 applications for non-traditional legal services provision. Similar initiatives are also under consideration in California, Florida, and Chicago.

In Canada, the Law Society of British Columbia made the first move by introducing an “Innovation Sandbox” last fall. It has already received over thirty applications and as of March 8 had approved five (only one application has been denied, and others remain under consideration). Applications received have included a virtual legal clinic, online wills services, and a program for legal research using artificial intelligence.[1]

II. Why are we doing this?

Enough has been said about the access to justice gap. I will not belabour that point. It is refreshing to see legal regulators try new approaches to address this gap, and a regulatory sandbox is a worthwhile approach for a few reasons.

First, legal tech providers currently operate in an environment of regulatory uncertainty. Let’s pause to recognize that legal tech is already providing some legal services. An August 2019 review identified 88 direct-to-public legal tech tools operating in Canada. Ontario-based legal tech companies are analyzing contracts, drafting legal research memos, and predicting case outcomes. At present, some legal tech companies (including the ones just linked) operate by providing their services to lawyers and law firms—rather than directly to consumers. While this business model surely has other benefits, one suspects it stems at least in part from regulatory requirements prohibiting non-licensees from providing legal services, as it ensures that consumers receiving their legal advice from regulated licensees with errors and omissions insurance, rather than from the (unregulated) tech company itself. But this model also risks doing little or nothing to narrow the access to justice gap, as consumers still need to retain an expensive lawyer to gain access to the innovative services.

It is not entirely clear whether the LSO considers the services currently provided by legal tech companies to fall within the broad meaning of the “provision of legal services” under the Law Society Act (“engag[ing] in conduct that involves the application of legal principles and legal judgment with regard to the circumstances or objectives of a person”). There is a risk these companies could be prosecuted for the unauthorized practice of law (or face other regulatory consequences) were they to provide services directly to the public. Indeed, the Technology Task Force’s report states that innovative technological legal services (or “ITLS”) “provided by persons not licensed as lawyers or paralegals may be subject to prosecution for unauthorized practice of law”—and acknowledges that “[t]his uncertainty can both deter the best innovation and deprive ITLS consumers of basic safeguards accorded to clients of licensees.

A sandbox eliminates the regulatory uncertainty legal tech companies currently face. But the LSO’s latter point is also important. Without any regulation of legal tech, it is not just the tech companies who bear some risk; consumers who use legal tech are left without the safeguards that legal regulation provides, namely, some assurance of competent and ethical services, and availability of recourse when needed.

One LSO bencher who objected to the sandbox was quoted as saying that “the sandbox regulatory framework will prevent any meaningful innovation”, suggesting that “the best way to kill innovation is to regulate it”. This objection misses the mark; the bencher failed to mention (or appreciate) that at present the regulatory framework prohibits the provision of legal services by tech companies entirely. The sandbox may not be a perfect solution (more on this below), but if the concern is stifling innovation, there can be no question that a sandbox is better than the status quo.

The second major benefit of a regulatory sandbox is the opportunity for data collection. Reporting requirements will permit the LSO to gather information about consumer demand for innovative legal tech services; how consumers’ sensitive information is being used and protected; what legal services can be provided to consumers by innovative providers in a more cost-effective manner to address unmet legal needs; consumer satisfaction with innovative non-licensee providers; and the extent to which any anticipated risks come to fruition (and how they might be effectively addressed).

Lawyers have a tendency to assert—without evidence—that the practice of law requires the education and skill of a fully-trained lawyer, and that clients cannot be assured their legal interests are being protected unless they receive legal services from a lawyer.[2] In some contexts, this may well be correct. But as emerging technologies become more prevalent and more sophisticated, it is difficult to believe that only lawyers can safely deliver all legal services. And the best way to assess this claim is by gathering and reviewing the evidence about alternatives.

III. Questions and concerns going forward

Ontario’s proposed sandbox (and, for that matter, the one in BC) differs from the one in Utah in a few ways. Although it is too soon to tell how Utah’s sandbox will turn out, I would suggest that Utah’s approach may be preferable on two points:

1. What happens after the sandbox?

In Ontario, it is explicitly left open “whether, and under what conditions, participants may receive a permit to continue providing the services after their participation in the sandbox has ended.” In Utah, providers admitted to the sandbox will be permitted to continue operating after the sandbox period has concluded unless the data collected as part of the project shows their services cause harm to consumers.

When identifying reasons to proceed with the sandbox, the LSO’s Task Force recognized that “Thoughtful developers and investors may not want to operate in an environment where they risk being shut down by the regulator or becoming a test case for unauthorized practice”. But Ontario’s proposed approach risks simply kicking that can down the road. Unlike in Utah, the LSO offers no assurances about whether, and under what conditions, sandbox participants will be able to continue beyond their two years in the sandbox; the proposal states only that “participants that have satisfied their performance objectives at the conclusion of the sandbox period may be given permission to continue operating in Ontario on an ongoing basis” [emphasis added, see pp. 14 & 22]. It seems this will be left to the LSO’s discretion down the line. One worries that the lack of assurances regarding post-sandbox operations creates a risk that some entrepreneurs may choose to direct their innovative efforts towards other industries, fearing that a viable and safe operation could be shut down if the LSO ultimately elects not to grant permits or make permanent regulatory changes.

2. Should lawyers oversee the regulation of innovative legal services providers?

Utah’s sandbox is operated by a newly-established “Office of Legal Services Innovation”. Utah made a conscious decision to have a new and distinct body—one that is not dominated by lawyers—oversee entities that will provide innovative legal services (which entities need not be owned or operated by lawyers). The Office of Legal Services Innovation is composed of “a mix of lawyers and other processionals… including staff with backgrounds in economics and data analysis to understand the impact on the legal services market”.

In Ontario, on the other hand, it is the Law Society—a body whose governance is dominated by lawyers—that will be responsible for the proposed sandbox.

This creates a potential conflict of interest: lawyers and paralegals, elected by lawyers and paralegals, all of whom currently have a sort of monopoly on the market for legal services, will be responsible for determining whether a new class of providers will be permitted to compete in the legal services market.

Of course, in Canada, our law societies have a statutory mandate to protect the public interest. But I would respectfully suggest that this doesn’t fully answer the concern; even if one assumes benchers and Law Society staff are acting in good faith and not engaged in anti-competitive behaviour, there remains a serious risk that they have a lawyer-centric bias about what is in the public interest.[3]

Moreover, the fact remains that benchers are elected by lawyers and paralegals (save for a small minority of government-appointed members). The licensees who vote for them may have anti-competitive tendencies, and there is a risk this could infect the decision-making process. In 2015, the issue of alternative business structures or “ABS” (i.e. where lawyers would not need to own 100% of a legal services business) dominated our bencher campaign, with many successful candidates campaigning on the basis that they would reject the ABS proposals in an LSO discussion paper of the day. Unsurprisingly, the Convocation that emerged did just that. One fears a future bencher campaign in which candidates take a stance against “robot lawyers” taking our clients or harming their interests—regardless of what the data collected in the sandbox might say about how innovative legal technologies can serve the public’s interests.

Another concern is that perhaps lawyers and paralegals are not the right persons to be regulating tech companies. Lawyers as a group are not exactly well-known for our technological prowess (to offer one example over and above everybody’s favourite cat lawyer, it took until January 2021 for Ontario’s court rules to start presumptively permitting service by email, having been stuck in the era of the fax machine up to that point). It does not necessarily follow that humans who are experts in providing legal services also have the expertise to govern innovative technologies that provide legal services in a different way.[4]

By creating a new body made up of lawyers as well as professionals with other specialized expertise, Utah has addressed both the conflict of interest and expertise concerns expressed above. The LSO (and LSBC), on the other hand, moved forward with a sandbox on the assumption that it is the right body to regulate legal tech. This assumption warrants closer examination down the line.

With all that being said, since beginning to regulate paralegals as well as lawyers in 2007, the LSO’s statutory mandate has not in fact been to regulate lawyers—it is to regulate “the provision of legal services”. Although this would not necessarily be the case in other provinces, significant statutory amendments would be required in Ontario to permit another body to regulate other legal service providers, a process that would surely be more time-consuming, controversial, and likely to get bogged down. If the government sought to step in and regulate some legal services, we would almost certainly hear concerns from the bar about the independence of the profession. Finally, as a practical matter, nobody else has offered well-developed proposals to fill the regulatory void for legal tech. Something had to change, and the LSO deserves credit for addressing the issue, potentially recognizing that even if it is not the ideal body to regulate new legal tech providers, it is the body best positioned to step up and move forward now, when there is little time to waste.

While Utah’s approach of building a new body to regulate the provision of innovative legal services in the 21st century shows promise, it is worth remembering that the LSO’s sandbox is a pilot project. It is intended to glean insights about what consumers need and how those needs can be safely and effectively addressed, and there is room to build responsive and principled regulation based on what is learned.

IV. Where might we go from here?

Now that Convocation has approved the regulatory sandbox, its development will move forward over the summer, with a target of October 2021 for accepting applications. I, for one, am optimistic about the sandbox and the insights it might provide for legal regulation going forward.

Assuming risks of harm to consumers can be satisfactorily addressed and it is determined that legal tech providers ought to be regulated in the long term, the LSO’s sandbox may take us in a few directions.

One option is that the LSO could continue as the regulator of “the provision of legal services”—whosoever provides them. At present, we now have two (main) classes of licence: L1 licences for individual lawyers and P1 licences for paralegals. Perhaps, in the not-too-distant future, we will see a new “T1” license for legal tech providers offered by the LSO.

Alternatively, it may be determined that while legal tech ought to be regulated, the Law Society as currently constituted (i.e. governed by lawyers and paralegals) is not ideally suited to do it, including for the reasons noted above. This would create an interesting decision point for the LSO. Should different classes of providers each have their own regulators, as is the case among regulated health professions? Should we revert to a Law Society that is the self-regulatory body for lawyers alone, and create a new body to regulate all other providers? Either of these two options would effectively reverse the 2007 decision to bring paralegals into the LSO’s regulatory scope (one notes that this was not the only option at that time; two reports commissioned in 1990 and in 2000 had recommended paralegals be regulated separately from the LSO, but this never came to fruition, including because of concerns that having different bodies responsible for regulating different aspects of legal services provision might lead to public confusion about who is responsible for what, or to inconsistency in standards and enforcement). If there is good reason to have just one regulator for all legal services providers, ought it be governed not only by lawyers and paralegals, but also by representatives of alternative providers (see a previous column by Malcolm Mercer in this space: “The Bencher from Amazon?”)? Put simply, it is not obvious what regulatory structure would best serve the objectives of legal services regulation going forward.

There are likely to be difficult decisions ahead. But, for now, I see reason for optimism: Canadian law societies are trying new approaches to promote innovation in the provision of legal services, and are seeking to do so in a way that prioritizes consumers’ interests and evidence-based decision-making.

Here’s hoping we take guidance from the evidence—and keep our focus on consumers’ interests—as we make the difficult regulatory decisions that will follow.



[2] This statement is paraphrased from the Toronto Lawyers’ Association’s response to the LSO’s call for comments on the 2016 report of Justice Annemarie Bonkalo recommending the expansion of paralegals’ scope of practice to include some family law services: Law Society of Ontario, Family Law Services Review Call for Input: Organization Submissions, p. 463.

[3] We’ve all heard the expression “to a hammer, everything looks like a nail”; I often suspect that, to a lawyer, every problem looks like it needs a lawyer. Noel Semple offers a much more sophisticated explanation of “lawyer-centricity” in his book Legal Services Regulation at a Crossroads: Justicia’s Legions (Edward Elgar, 2015) at s. 3.1 (beginning on p. 112).

[4] This may be at least partially addressed by the Task Force’s proposal to establish a volunteer Advisory Council to “help steer the sandbox to meet its objectives”, whose “members would represent a range of expertise including: legal technology and innovation; legal regulation and professional ethics; priority legal practice areas such as family law; consumer protection and advocacy; economics; regulatory sandboxes and government or judicial administration.” The Task Force notes that this would “allow the regulator to tap into skills and perspectives that it lacks in-house” and give the public confidence that the regulator is “open to exploring new ideas, guided by leading experts”. Indeed, if it proves to be useful to have outside experts advise in this context, the Law Society may glean insights for its regulatory model more generally!


  1. Thanks for giving us a good overview and thoughts on the LSO sandbox project! I was a bit fuzzy about what it was about.

    This part – “Another concern is that perhaps lawyers and paralegals are not the right persons to be regulating tech companies.” – I’m fully in agreement. I think some external thinking and perspective are mandatory to move legal practice ahead at a reasonable pace.

    I’m curious and excited to see what applications emerge. At the very least, I hope it gives the professions some insights into technologies and services that we never really contemplated but that may have real value to the public.

  2. Great article. It will be interesting to see what emerges.