Agentic AI

Verification infrastructure for autonomous systems

An autonomous agent acts across a system or organisational boundary. Before the receiving system accepts the action, it can independently verify that the agent is still authorised and within its limits.

Authority granted

An autonomous agent is authorised to act within defined limits.

Within the agent’s organisation, a principal grants authority to an autonomous agent. That authority defines what the agent may do, the limits that apply and how long it lasts. This illustration shows authority being granted, before any receiving organisation’s decision about an action.

Defined authority

Principal: Grants the agent its authority.
Scope: What the agent is allowed to do.
Limits: The limits that apply to the agent’s actions.
Duration: How long that authority lasts.

Authority can change before an action happens

Permission at the start does not establish what an agent is allowed to do later. Some authority may already have been used or committed. It may also have expired, been revoked or been reduced.

Guardrails help govern an agent’s behaviour, but they do not by themselves establish its current authority for a particular action.

The question at the point of action is whether the agent’s authority is still valid and sufficient for that request.

plusChain™ for Agentic AI

Patent pending

plusChain provides verification infrastructure for actions across systems and organisations. It enables the receiving side to independently determine the agent’s current authority before accepting an action.

Check current authority

Consider what was granted, what has been used or committed, and changes to scope, limits, duration or policy.

Verify independently

The receiving system determines current authority from relevant evidence, independently of assertions from the agent or the organisation operating it.

Keep delegation within its limits

A sub-agent receives narrower authority that remains dependent on the parent’s authority. A change to the parent’s authority can change what the sub-agent is permitted to do.

plusChain supplies the verification infrastructure. The receiving organisation retains the decision about whether to accept the action.

Explore Agentic AI

Potential application areas

Examples of authority questions across different settings:

Procurement and spending

Is the agent still authorised to make this commitment within the applicable limit?

API and tool access

Is the requested operation within the agent’s current authority?

Data access

Does the agent still have authority to access the requested data?

Resource allocation

Does the allocation fall within the agent’s remaining authority?

How it works

Follow an action from the authority granted to an agent through to a receiving organisation’s decision.

Authority is established
An agent is authorised to act within defined scope, limits and duration.
An action is requested across a boundary
The agent requests an action in another system or organisation.
Current authority is verified
The receiving side independently checks whether authority is valid and sufficient. A completed check can find that authority is insufficient.
The receiving organisation decides
An action is withheld when authority is invalid or insufficient. The receiving organisation retains the business decision even when authority is sufficient.

See how it works

plusChain for Digital Assets

Patent pending

A digital token’s history does not, by itself, establish the provenance of its underlying asset. plusChain for Digital Assets is designed to help check the recorded connection between tokens and their underlying assets, including recorded authorship and transaction history, and identify changes or inconsistencies in the associated records. The invention addresses provenance across multiple marketplaces or blockchains. Its focus is asset and token provenance, distinct from the agent-authority verification addressed by plusChain for Agentic AI.

Explore Digital Assets

The team behind plusChain

Our co-founders bring experience in enterprise software, public-sector procurement, investment banking and fund management.

Meet the team

Discuss your authority-verification requirements

Tell us where autonomous agents act across your systems or organisations, and what the receiving side needs to verify.