Trusted & Certified Digital Identity Must Be the Foundation of Agentic AI
Agentic AI will transform services across the UK economy by enabling systems to complete tasks and transactions on behalf of people and organisations, rather than simply assisting them. Combined with trusted digital identity infrastructure, agentic AI could significantly reduce friction across onboarding, payments, procurement, healthcare administration and regulated public services by enabling verified AI agents to securely act on behalf of individuals and organisations at scale.
Experts predict digital identity systems will unlock economic value equivalent to several percentage points of GDP when deployed across high-value digital transactions and services at scale. As AI agents increasingly conduct trusted transactions and service interactions autonomously across sectors, realising this opportunity will depend on trusted digital identity, delegated authority and interoperable trust frameworks being embedded into the infrastructure layer from the outset.
The UK Digital Verification Services (DVS) trust framework provides an important foundation for this by enabling interoperable, reusable and trusted digital identity across AI-mediated transactions and services. Trusted and interoperable digital identity infrastructure could become a foundational layer for trusted AI-enabled economic activity across the UK.
Key opportunities include:
onboarding, fraud prevention and payments in financial services,
healthcare administration and patient services,
public services and benefits delivery,
bookings, payments and age-restricted purchases,
recruitment and right-to-work verification,
and procurement and supply chain management.
However, large-scale adoption of agentic AI will depend on confidence, accountability and interoperability. The organisations best placed to solve this are those who have been thinking seriously about trust, privacy and certification for identity solutions.
As AI systems begin acting autonomously on behalf of people and organisations, there must be assurance regarding:
who or what is acting,
what authority has been delegated,
what data can be accessed,
and who remains accountable for decisions and actions.
This shift towards AI agents acting as authorised transactional actors across the economy will require emerging “Know Your Agent” (KYA) models, reusable digital identity credentials and continuous trust evaluation to enable safe and scalable deployment.
Without interoperable identity and delegated authority infrastructure, organisations will struggle to safely scale AI-driven transactions and services across sectors and platforms.
The Role of the UK DVS Trust Framework
The UK DVS trust framework establishes common rules and standards for digital identity and attribute verification services. It enables organisations to rely on digital identities and credentials that have been created, verified and managed to consistent standards.
Following the Data (Use and Access) Act, the DVS framework is now on a statutory footing in the UK, creating an important opportunity to establish trusted digital identity as core infrastructure supporting AI and the wider AI-enabled economy.
The DVS framework supports:
authentication,
consent,
delegated authority,
credential verification,
audit trails,
accountability,
secure digital credentials,
interoperable digital wallets,
and reusable digital identity credentials.
Reusable digital credentials can help create portable trust across platforms, sectors and AI-mediated services. These capabilities are essential if agentic AI systems are to operate safely and at scale across the economy.
Agentic AI Changes the Trust Model
Traditional digital interactions are based on a direct relationship between a user and a service. Agentic AI changes that model by allowing individuals and organisations to delegate authority to AI agents operating across multiple systems and services.
AI agents may increasingly:
open bank accounts,
authorise payments,
manage healthcare appointments,
complete age assurance checks,
interact with regulated public services,
negotiate procurement processes,
or access reusable digital credentials and delegated permissions across platforms and services.
In these environments, systems must be able to determine:
who the agent represents,
whether authority has been delegated,
whether permissions remain valid,
and whether actions remain within authorised limits.
Without interoperable and trusted digital identity systems, these environments quickly become difficult to manage safely.
Secure Binding and Delegated Authority
A critical requirement for trusted agentic AI is secure binding between AI agents and verified real-world identities.
Where an AI agent acts on behalf of an individual or organisation, there must be strong assurance regarding:
who authorised the agent,
what authority has been delegated,
what actions are permitted,
and how permissions can be revoked or limited.
This cannot rely on weak authentication methods such as one-time passwords or email-only verification.
Regulated environments require:
DVS-certified digital identities,
reusable digital credentials,
biometric verification with iBeta liveness detection level 3 to ‘bind’ the correct human
and cryptographic proof mechanisms.
Systems must be independently audited to DVS trust framework certification levels and also be able to demonstrate that:
authority has been explicitly delegated,
actions are uniquely bound to verified identities,
activity remains auditable,
and delegated permissions can be revoked where necessary.
Auditability and Accountability
Strong auditability will be essential for organisations, regulators and consumers.
Systems must maintain records showing:
which agent acted,
who authorised it,
what permissions were granted,
what actions were taken,
and when those actions occurred.
This will be critical for fraud prevention, dispute resolution, regulatory compliance, liability management and public confidence.
Government and Regulation Have a Critical Role
Government and regulators can support responsible adoption of agentic AI by:
clarifying legal acceptance of trusted and DVS certified digital identity across sectors,
establishing standards for secure identity binding and delegated authority,
supporting interoperable and privacy-preserving DVS identity solutions,
clarifying liability and accountability rules,
and enabling trusted digital credential ecosystems.
The UK also has an opportunity to establish globally trusted standards for AI-mediated digital identity, reusable credentials and delegated authority, helping position the UK as a leader in trusted agentic AI infrastructure.
Building Trusted Agentic AI
Trusted digital identity under the UK DVS framework can provide the trust layer required for safe and scalable agentic AI and the wider AI-enabled economy.
However, trusted deployment depends on more than authentication alone. AI agents must be securely bound to verified individuals and organisations, with clear delegation of authority, strong auditability and robust accountability built into system design from the outset.
The UK has an opportunity to lead globally and realise the economic and societal benefits of trusted agentic AI. But this will only happen if legislation, regulation, standards and industry practices clearly support trusted and certified digital identity, secure delegated authority and accountable AI systems across sectors. Yoti has been building these foundations for trusted digital identity since 2014 and we stand ready to support businesses on this journey.
