FB pixel

Daon expands AI agent governance patent portfolio

Latest patent adds action-level authorization to the company's emerging AI agent governance framework
Daon expands AI agent governance patent portfolio
 

Daon has patented a system that combines human identity, agent behavior, and runtime context before granting an autonomous AI agent permission to call a tool or an API.

U.S. Patent No. 12,688,261 describes the authorization process that checks whether the agent is still linked to the intended person, if the agent is behaving as expected, and the context of the request, including the tool, action, resource, scope, session, and execution environment.

If the authorization request meets the required conditions, the system generates a machine-verifiable “delegation artifact” that defines the permitted action, scope, and validity period.

The machine-verifiable entity could be a token, a credential, a signed assertion, or a cryptographically protected authorization object.

The permission can be configured for the type of task. In a financial workflow, it could limit the value of a transaction. In other settings, it could limit how much data the agent may access, how often it may use a tool, or whether it may perform restricted actions such as changing account settings or deleting records.

This shows that the result of the authorization process is not limited to a binary allow-or-deny decision. If the system detects greater risk, it can restrict what the agent may access, reduce how long the permission remains valid, or limit how often the agent can act.

The patent is Daon’s third covering AI agent governance.

The earlier patents addressed the agent’s continuing connection to a person and the integrity of its runtime behavior. Daon says the three patents create a progression from person-agent fidelity to behavioral integrity and action-level authorization.

The patent portfolio also extends Daon’s work on AI governance. The company recently received ISO/IEC 42001 certification following an independent audit of its management system governing the AI lifecycle.

The described patented capabilities are not commercially available.

Related Posts

Article Topics

 |   |   |   |   |   | 

Latest Biometrics News

 

Digital and trust infrastructure investments unlock value of identity

As the role of digital identity in society evolves and biometrics increasingly form the foundation of trust in remote interactions,…

 

Listen: How age assurance platforms earn trust

The world has more or less decided that some online services should be age-restricted. The question now is, how to…

 

Sri Lanka sets Q4 target for SL-UDI rollout

Sri Lanka is preparing to issue its first production digital identities next quarter, beginning a phased rollout of the country’s…

 

Madagascar strengthens trust architecture for national digital ID

Madagascar is strengthening the legal, institutional and technical foundations of its national digital identity program, introducing new cybercrime legislation as…

 

Cybastion backs Cameroon digital sovereignty with $75M infrastructure project

Cybastion has committed $75 million to fund the construction of digital infrastructure in Cameroon, including a data center and an…

 

Co-Develop, Caribou outline governance blueprint for DPI adoption

Digital public infrastructure (DPI) programs have grown to include more than 100 countries globally. Recognizing this expansion, Co-Develop and Caribou…

Comments

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Market Intelligence

Featured Company

Biometric Update Podcast

Most Read This Week

White Papers

Latest Webinars

Biometrics Industry Events