Unleashing the potential of Verified Attributes

What is the one, single thing the UK Government could do to boost the operation and performance of the digital economy? In this blog, we explore the massive untapped potential of Verified Attributes.

What is a verified attribute?

A verified attribute is any piece of information about a person or performance that has been checked by a responsible trustworthy body and made available to another party.

A driving licence provides multiple verified attributes including the holder’s:

• Age / date of birth
• Address
• Other aspects of their identity including a picture of their face and their signature
• Verification that they have indeed passed their driving test for one or more classes of vehicle.

The issuing body in the UK, the DVLA, is widely acknowledged as having robust processes to make sure it only issues driving licences to people who entitled to one. One positive unintended consequence of DVLA-issued driving licences is that it has also created a mechanism for trusted data sharing. In addition to its formal use as a verification of a right to drive, driving licences are used in many different ways, for example:

• By young people wanting to prove they are old enough to buy alcohol or enter a club
• Presenting to a post office to pick up a parcel as proof of identity
• Applying for a job where having a valid driving licence is an essential criterion for eligibility

The verified attributes contained in driving licences have been widely used for multiple purposes for many years. They are not underwritten by the Government or DVLA with any liability model, but it has nevertheless been good enough for relying parties to accept. So why not make such data portable as a collection of digital verified attributes that the citizen can carry with them? In a digital economy, the driving licence model can be applied and dramatically extended by responsible bodies issuing secure electronic tokens that encapsulate many different bits of information previously embedded in paper or plastic documents. Once an individual has been provided with such a token they can present it to any other ‘relying party’ that wishes to have the information in question confirmed.

For example, if the Department of Work and Pensions (DWP) issued a token to an individual confirming that they are on one or more benefits, this individual could use this token to apply for Warm Home Discounts from energy suppliers. By issuing this token, the DWP would be helping both individuals and energy suppliers eliminate administrative hassle and cost and streamline the process for all concerned, including DWP. One small change but a massive beneficial impact for everyone.

There is an almost infinite variety of attributes that could be verified by responsible bodies. They include:

• Personal attributes such as date of birth
• Qualifications: confirmation that the individual concerned has passed certain examinations, is qualified to driving a car, heavy goods vehicle or pilot a plane etc.
• Eligibility: the fact that the individual has passed certain tests, e.g. that they have a certain score for credit worthiness; that they are ‘officially’ disabled and eligible for disability allowances; that they are entitled to a warm home discount
• Personal history: for example, verification that the individual was indeed employed by a particular employer between certain dates; paid their taxes; legal and other judgements
• Other attributes: for example, in the world of medicine, many different attributes could be provided by health service providers including information about blood type, allergies, vaccinations, congenital diseases, operations, treatment entitlement etc.

Paper and plastic verified attributes are currently used in millions of transactions every day, including proofs of entitlement, applications (for jobs, loans etc), and identity verification. But they haven’t gone digital. Yet.

The problem

If the sharing of verified attributes via secure electronic tokens could bring such huge, immediate benefits to our economy and society, why isn’t it already happening?

There is a simple answer: verified attributes are a classic collective action problem resulting in market failure. While all individuals and almost all organisations would benefit greatly if a wide range of verified attributes were available for quick, easy, safe sharing and checking there is no immediate particular benefit to any single attribute provider to incur the costs of providing citizens/customers with these verified attributes about themselves.

For example, DVLA does not get any immediate benefit from providing driving licence holders with secure digital tokens verifying their age, address, etc. Indeed, it could argue that incurring costs on such activities might be ultra vires, because it has nothing to do with its core purpose as an organisation.

But data portability rights under GDPR change all that. Or could do, anyway, because they give individuals a right to a copy of their data.

Implementing a portable Verified Attribute policy

How could/should the Government implement a portable Verified Attribute policy? GDPR/Data Protection Act offers two broad ways of implementing data portability:

1. Organisation-to-organisation data sharing, where permission for data sharing for clearly defined purpose is granted by an individual
2. Data controller (organisation) to individual, where the individual asks for a copy of the data to be ported to the individual’s own storage device or service for their own personal future use.

In our earlier Briefing Paper on data portability we argued that this second option avoids many problems and opens up many opportunities, especially if implemented by providing individuals with secure Personal Data Stores that help them receive, store, curate, manage and share their own data safely and efficiently.

In theory, a policy of portable Verified Attribute sharing could be implemented in both the above ways. However, with portable Verified Attributes, the disadvantages of the first option are magnified as are the advantages of the second option.

If implemented in an organisation-to-organisation way, some organisations such as the DVLA or Passport Office could be inundated with multiple separate requests for Verified Attributes from multiple parties. This could place a significant administrative burden on these organisations, while also creating cyber security threats and horrendous complexity around informed consent. Organisations would have to build data sharing relationships with other organisations that have nothing to do with their business, checking their identities before sharing any data with them and exposing their large-scale back-end systems to the outside world.

However if the Verified Attribute is given directly to the individual via that individual’s Personal Data Store (PDS) (and if it is maintained thereafter to ensure they always have an up to date current record), organisations only have to provide the Verified Attribute once (and maintain updates to it). Any time individuals want to do anything with that data, they are free to onward-share the token when and if they need to.

This is more secure, easier to manage and lower cost in terms of processing and capacity. It simply extends the organisation’s existing relationship with the customer becoming part of its service to that individual, and avoids forcing the organisation to deal with multiple other organisations which it has no prior relationship with. On top of that individuals are incrementally empowered with their own data, building up their own rich library of Verified Maintained Attributes over time — a library of attributes that they can use for multiple purposes to manage their lives better. (One of these purposes would be to prove their identities when dealing with other organisations — a massive issue in its own right which we explore in detail in our next blog.)

To keep such a system of Verified Attributes running effectively, once an attribute is shared, it needs to be maintained via a persistent link to the individual’s PDS so that any status changes are instantly and automatically relayed to the token held in the PDS. (For example, if an individual claims to have a driving licence but has recently been banned from driving, if they are asked to share a token verifying this fact and refuse to do so, this creates an alert prompting further investigation.) But in all other cases (where the individual has not been banned or is happy to share this fact) the data can be shared instantly, automatically, safely and cheaply with the attribute user knowing they can trust the data to be true and up-to-date.

The up-to-date maintained status of Verified Attributes is vital to parties relying on the accuracy of the data: if they are not confident the data is up to date and accurate they have to start making checks, which introduces new (perhaps manual) processes which undermine the whole point of the enterprise.

Verified attributes and data portability

To sum up. Huge personal, social and economic benefits are currently being left neglected because of a damaging collective action problem and market failure around the sharing of Verified Attributes. This could be resolved quickly and simply by the relevant authorities by:

1. Officially interpreting Clause 20 of GDPR/Data Protection Act to include verified attributes, so that if a customer asks a service provider to provide a verified attribute in a secure electronic form the data controller is legally required to do so.
2. Encouraging public sector service providers including (in the UK) DVLA, Passport Office, HMRC, DWP to proactively offer their customers access to a certain number of pre-agreed verified attributes, so that they can use them anywhere they need them.

The time is ripe

If the Government pursued a policy of Verified Attribution sharing via Personal Data Stores, it would:

• immediately remove a huge amount of cost, friction, administrative hassle and risk from billions of day-to-day transactions including applications, provision of proofs of entitlement and so on, for both service-providing organisations and citizens.
• go a long way to solving ongoing problems relating to identity assurance (see our next blog)
• create the infrastructure and mechanisms necessary for providing truly joined-up services in the public, private and third sectors in a highly efficient manner, in a way that protects individuals’ data and privacy (because they control which attributes are shared with who)
• begin to unleash the true power of data for service innovation including automation and AI able to work using trusted data.

What we have outlined is just common sense, part of what’s necessary to build a safe, efficient, value-creating personal data infrastructure. Fixing Verified Attributes would kill many birds with one stone. The opportunities are immense. Many in Government now realise this.

The time is ripe for what could be one of the most important moves towards the creation of a 21st century digital economy, fulfilling the potential envisioned in the UK industrial strategy, the Health and Social Care, Civil Society strategy and delivering competitive functioning markets.