Trust is the foundation for AI-agents

By now it is clear - not only to the ChatGPTs I have asked (and posted here) - but also to most of us - that the identity building credentials in general purpose organisational wallets are a must-have when employing AI-agents. The bicycle made for two (personal identity used at work and organisation's wallet) can then safely be equipped with an AI-agent.

A very good description of the need for trust found here: https://www.linkedin.com/posts/mikahuhtamaki_digital-trust-starts-with-identity-activity-7313086083095834624-Vc4C?utm_source=share&utm_medium=member_desktop&rcm=ACoAAABPj1oB9_D7YNYACmHvY9HioUqpuULqZCo 

Markus Hautala commenting: 
"As you mentioned in your article, AI agents are not entirely autonomous entities in the sense that only natural persons and legal entities can enter into legally binding agreements. While AI agents may, in the future, automate tasks on behalf of natural persons and legal entities, the ultimate responsibility for their transactions lies with these human and legal entities. Currently, a significant challenge for AI agents is their lack of deterministic behavior..."

"Regarding digital identity, such systems are used to model and represent the trust hierarchy within a specific digital trust ecosystem. This hierarchy is based on real-world trust arrangements, where various parties enter into agreements with one another, and these agreements are ultimately anchored in law to ensure their legal validity.

Therefore, to enable AI agents to act on behalf of natural persons and legal entities in the future, we must be able to express the trust hierarchy between natural persons, legal entities, and AI agents using the tools and methods provided by digital identity systems."