We were pleased to be invited to participate in the World Economic Forum workshop last week on Cybercrime 2025 focusing on Digital Identity.
One of the participants presenting were Access Now, who “defends
and extends the digital rights of users at risk around the world” [https://www.accessnow.org]
who are running a campaign that I’d urge you to add your signature of support
for; called #WhyID.
They ask that at the onset of any digital identity programme
in any given region or country the #WhyID question must be asked;
Given that our aspiration is a global identity ecosystem,
then I guess responding to these questions are even more important for us as an organisation. So here goes;
1. Respond to WhyID?:
• Why do we need these foundational digital
identity systems? What are their benefits?
We need foundational digital identity as we live in an increasingly digital
world that has little trust, and in a digital world where the majority of
entities are based on self-asserted identity.
In short, the benefits, if we do this correctly, are;
o
The move from self-asserted identity and
identity attributes, to trusted identities with attributes from truly authoritative
sources.
o
The move from identities that operate only
within a locus-of-control, to identities that can be reused anywhere, by anyone,
globally.
o
The move from identities that need to have a
central authority at its heart, to a decentralised, privacy enhancing ecosystem
[and one that is NOT blockchain][1].
o
The move from a binary level of trust, to one
where the entity taking the risk (remembering that risk is bi-directional, yet asymmetric)
is able to understand the risk of every component part.
o
The elimination of billions of dollars of fraud
and crime.
o
The elimination of identity theft and impersonation.
o
The ability to understand information from
trusted, traceable and reputable sources, vs. un-trusted, self-asserted and fraudulent
entities (trolls, sock-puppets, state sponsored misinformation etc.).
o
The ability to leverage global ecosystem for
secure and trusted IoT devices and secure and trusted communications.
• Why
are such programmes deployed without sufficient evidence of the benefits that
they should deliver? How do these programmes plan to reduce the risk to and
safeguard the rights and data of users?
We agree; most programs are designed to only fix one particular issue,
and are limited in both scope and thus design.
In contrast, we started by looking at why Identity systems fail[2],
from there developing this understanding of what you need to “do differently”
to build a set of principles[3],
and from there designing a system[4]
to meet those principles.
Thus the model builds in privacy by design, ensuring anonymity where needed and
places the identity of an individual entity under the full control of said
entity, with no intermediate systems or infrastructure that can be compromised.
• Why should it be mandatory – either
explicitly or de facto – for users to enrol onto these programmes? These
programmes are either mandatory through legislative mandates or through making
them a precondition to essential services for users.
We feel it should not; an entity should be able to generate its own root with
100% anonymity, and with total control over that root. Said entity should be
able to generate personas (the join of said entity and an entity that is authoritative
for one facet of said entities overall identity) only when there is a
benefit to said entity [you only need a passport because you want to travel
across borders that require passports].
Most entities will see the benefit, especially as the use of a common (cryptographic)
root [albeit 100% anonymous] allows multiple privacy enhancing assertions to be
made from disparate personas as a provably linked set [only the one entity
could have made them]. For example: “I am over 21” & “Here is payment for alcohol”.
• Why are these programmes centralised and
ubiquitous? Why is one digital identity linked to multiple facets of a
citizen’s life?
We see this as one of the fundamental questions; and our stance is that designing
a digital identity system in this manner is fundamentally wrong, technically unnecessary
and ultimately causes any digital ecosystem to fail or implode.
While there a great benefits to having multiple, disparate, trusted attributes
all under a central “root” (after all, this is what happens in real-life); you
can only make this work if that root is 100% anonymous; the design must also
take into account when the entity in question decides their level of trust in
the ecosystem is insufficient and allow them to have multiple, unconnected roots.
• Why
are countries leapfrogging to digital identity programmes, especially in
regions where conventional identity programmes have not worked? The scalability
of digital identity programmes also makes their harms scalable.
We believe (based on historical evidence) that identity ecosystems implemented
at a national level either fail, implode to a sub-set of services and fail to
federate (be trusted) outside of that particular locus-of-control.
Instead, giving away for free, an eco-system and a standard that needs no
central infrastructure; which is therefore simple to adopt; where the government
or organisation is only responsible for its people and only for those attributes
for which it is truly authoritative delivers all the benefits to countries and
their citizens without the potential harms that come when such a system is scaled.
• Why are these digital identity programmes
not following the security guidance coming out of various expert academic and
technical standard-setting bodies on the use of biometrics in identity systems?
We’d go further than this and suggest that any biometric used for
authentication should never be stored by any third-party.
This does not of course preclude the nefarious collection of biometric
information (fingerprint from a glass) or the (legal, or illegal) use of biometric
recognition systems (typically facial or gait) linked to surveillance systems.
Instead, a digital identity ecosystem must be designed to understand the
authentication of the entity to the digital as well as the level of trust it
can place in an assertion of biometric authentication (but not validating the raw
biometric itself) and in such a manner as to render replay attacks useless.
• Why are some private sector enterprises
being privileged with access and ability to access the ID systems and build
their private businesses on top of them? What safeguards are being implemented
to prevent the misuse of information by the private sector? What should be the
role of the private sector in the identity ecosystem?
The driver for most companies is the ability to make money; either from
building large identity infrastructure (either traditional or more recently
blockchain), in the form of consultancy or through controlling access to
attributes.
Instead we believe that no big infrastructure is required; and organisations
that are authoritative for facets of an entities identity must be able to add
the necessary service to their existing systems to be able to sign trusted attributes
that can be held, maintained and managed by the entity to which they pertain.
In addition, organisations wishing to consume said trusted, authoritative attributes
when proffered them by said entity, must be able to add the necessary service
to their existing system to accept and validate these.
We envisage both add-on's being open-source and royalty free to ensure proper
security validation and widespread global take-up.
Those who promote these programmes must first critically evaluate and answer these basic WhyID questions, along with providing evidence of such rationale. In addition to answering these questions, these actors must actively engage and consult all actors. If there is no compelling rationale, evidence-based policy plan, and measures to avoid and repair harms, there should be no digital identity programme rolled out.
2. Evaluate and, if needed, halt: The potential impact on human rights of all existing and potential digital identity programmes must be independently evaluated. They must be checked for necessary safeguards and detailed audit reports must be made public, for scrutiny. If the necessary safeguards are not in place, the digital identity programmes must be halted.
We would agree; (and probably go further) as we believe that
adopting the Identity 3.0 principles[3]
and the associated global ecosystem will both protect human rights and provide greater benefits for the government and its citizens.
Our belief is that your biometrics (as they relate to authenticating your identity) should be collected, stored and validated under your direct and exclusive control.
Any relying entity wanting to validate the level to which an entity is authenticated should, along with the relevant signed attributes, be able to understand everything about how authentication was achieved (device, version, pass threshold, number of attempts etc.) allowing them to make their own risk assessment of whether that is adequate for them, of course with the option to then use some form of “step-up” authentication should the biometric threshold be inadequate.
This way there can be no collection, and thus no breach, of an entity’s biometric information.