What is needed for digital regulation to succeed?

Digital regulation refers to a scenario where computers are able to communicate and process compliance data without requiring humans to get involved. Examples of digital regulations might include submitting your taxes electronically, although there are more interesting examples that we will explore.

Here’s our checklist for what we believe a system of digital regulation needs to be successful:

Rules must be freely available

We believe that it is imperative that users are able to download the rules and use them for their own purposes. The ability to download computer-readable rules means that they only have to be implemented once and that everyone can benefit from their use.

Rules should be useful

We also believe that the rules should be as useful as possible to realise the full benefit of digital regulation. More specifically, we believe that it should be easy to download the rules and integrate them with existing systems.

As an example, we don’t believe that it would be ’enough’ for the Government to provide an API where the public can upload all of their data to check it for compliance. There are multiple issues with this scenario (privacy comes to mind), as well as the impracticality of sending ‘all of the data’ over the internet, possibly many times per day.

Instead, we believe that the public should be able to download the rules and integrate them into them into existing systems. And while this is not a significant technical challenge on its own, many Rules-as-Code systems falter by not planning for this feature from the outset.

Rules must be trustworthy

This seems like an obvious requirement, but it is more difficult to imagine what it takes for rules to be ’trustworthy’. In our experience, the single greatest contributor to trustworthiness is readability; you have to understand what the rules are doing before you can trust them.

The next critical element is that the rules need to be tested against all expected scenarios and shown to produce the expected outcomes.

Non-developers should be able to write and review the rules

This is less of a requirement an more of a big indicator of success. Before an authority will be willing to sign-off on a digital regulation system, they will need to feel confident that they understand how the system will behave.

The more readable the rules are, the more trustworthy they become. Software is often delivered as a black box and a promise of, ’trust us’, which is not good enough for government.

Regulators will need to adapt as well

Moving to a digital-first approach will require regulators to stop thinking of compliance as a true/false proposition. Digital processes will (hopefully) uncover innocent mistakes that are made by humans, and regulators will need to be flexible in their response.

We believe that the majority of compliance violations are made by mistakes, primarily due to the inefficiencies of our current regulatory system. It would be foolish for the Government to immediately enforce every violation that the new approach uncovered.

And possibly more confronting is that Regulators will need to be willing to accept and admit that they will make mistakes as well. Specifically, it is possible (or even likely) that there will be some mistakes in the rules. Rather than seeing this as the reason not to embrace digital regulation, it is important that this is viewed as a positive attribute; mistakes can be caught quickly and updates can be efficiently distributed.