The sorry state of digital regulation

By Rupert D.E. Brown, Evidology Systems 

It is now four years since the FCA conducted its first “tech sprint” which tried to evaluate whether financial regulatory reporting requirements could be expressed in a formal specification language which could then be consumed in a systematic and largely automated fashion by regulated firms.

There have been several iterations and feedback cycles since then, but no concrete outcomes or a definitive implementation roadmap have emerged. It’s clear from reading some of the documentation that Brexit was also a factor in muddying the waters, especially when considering the jurisdictional scope of any programme.

It is also notable that a lot of participants focussed their feedback on a range of “magic” technology solutions, hoping that the “touchstones” of AI, NLP, Blockchain, and various semantic standards would provide a clear path to any solution.

All the statements they made were basic references to these “techno fetish totems” rather than any detailed explanations or designs which showed how these technologies would be used and whether they would work.

In this time of enforced isolation and reflection, perhaps we should take a cold, hard look at the commercial successes of these technologies, both in terms of physical footprint and balance sheet bottom line, coupled with traded stock price.

Formal computer programming and related data specification languages have now existed for over 60 years and, arguably, twice as long, when you look back to the Hollerith punched card data management solutions which started in the early 1900s.

One must therefore ask why we have not solved this comparatively simple reporting specification problem with such a mature set of capabilities and large practitioner base.

The FCA’s initiative is largely forward looking – what has never been contemplated, let alone discussed in any detail, is how we will bring existing statutes up to date.

As an example, take a look at the EU’s Payment Services Directive PSD2 – you can download a PDF version in your preferred national language from the EU’s LEX website.

Once downloaded, search for the term “Chapter” within it and you will soon find multiple instances of the same “Chapter” number, which relate to completely different sections (“Titles”) with the full text of the statute.

PSD2 is of course a comparatively modern regulation, and a simple hierarchic naming and numbering system could rapidly fix this ambiguous classification problem.

One might ask whether we need to fix this problem, and we do have a choice:

  1. Put the rock we have lifted back down on the problem and maintain the status quo for the definition and documentation of statute laws and many standards.
  2. Start to address this problem with a more precise form of unambiguous classification scheme for new documents as well as revising the existing statute.

Since the turn of the current millennium the world has been through many legal and commercial shocks, driven largely by much more tightly-coupled global communications mechanisms and supply chains.

Rupert D.E. Brown

The upshot of this is both broader and more detailed regulatory requirements. The current Covid-19 pandemic will add yet another layer to this, around resilience, biosecurity, and general hygiene.

We must therefore naturally choose option b), if we are going to draft and implement legal requirements at scale and with precision, in meaningful timeframes.

Choosing option b) will naturally cause significant discomfort to many in the legal profession who have relied on ambiguity and imprecision to ply their trade.

It can, however, be argued that neither the State nor businesses are prepared to tolerate these costs any longer, and have starved the legal sector whenever possible in recent years to the extent that it is no longer a leading career choice for graduating students.

Better tooling and automation must transform this profession, in the same way they did for design, manufacturing, media and finance over the past 40 years, since the widespread adoption of mobile and distributed systems technologies.

Over the past decade we have seen the slogans of FinTech and RegTech emerge as rallying points for technology vendors; the phrase LegalTech is now starting to emerge.

However, we need to be concerned that current definitions of the sector’s offerings amount to little more than a rebranding of document management, workflow and search, with a bit of the pixie dust of crowdfunding (which in some cases is, rather oddly, defined as LegalTech).

It is easy to resort to phrases such as “sorting out” or “getting a grip”, especially in these uncertain times, but like many things that the Covid-19 pandemic has highlighted, there is a clear and unambiguous need to rationalise the way statutes and standards are drafted and accessed.  Without change we will continue to wade in the muddy waters of imprecision and inefficiency.