Indicio joins Nvidia Inception Program to develop agentic AI orchestration

Indicio is planning to develop its digital identity infrastructure for agentic AI, with help from Nvidia. The Seattle-based firm has announced that it is joining the Nvidia Inception Program, a free accelerator program designed to support startups working with AI.

A release says Indicio’s technology, called ProvenAI – launched in July, riding on a wave of new agentic AI applications – enables both users and AI agents to authenticate and share data using decentralized identifiers and Verifiable Credentials (VCs). This means an AI agent can cryptographically prove who it is interacting with, and vice versa, before data is shared. VCs also authenticate the data that is being shared.

Once verified, the two trusted entities are free to transact and exploit the potential in hordes of AI agents. The technology promises to automate workflows for both companies and individuals, ushering in a workforce where humans and bots work in tandem, and in compliance with data privacy laws.

The critical benefit to using VCs for this task is that there is no need for either party to phone home to crosscheck a database during authentication or authorization. Because a Verifiable Credential is digitally signed, the original credential issuer can be verified without having to contact the issuer. The information in the credential can also be cryptographically checked to see if it has been altered.

GDPR concerns over consent, disclosure covered with automation

The biggest issue related to AI agents is security, as they need to be able to access sensitive data in order to execute tasks. “For AI agents to be useful, they must be able to access personal data – lots of it,” says Indicio in a blog. “For this to be compliant with data privacy regulations such as GDPR, a person must be able to consent to share their data. There’s just no way of getting around this.”

In terms of consent and selective disclosure, Verifiable Credentials fulfill key requirements of the GDPR.

Indicio’s software tells each participant in a network who is a trusted issuer and verifier, and which information needs to be presented for which use case in what order. But Indicio also wants to achieve something even more complex than regulatory compliance: a decentralized and automated way to govern VCs.

Decentralized governance (DEGov) allows organizations to automate and quickly create and implement the rules that Verifiable Credentials and their holders will follow.

“With Indicio Proven and ProvenAI, a network is governed by machine-readable files sent to the software of each participant in the network (i.e., the credential issuer, holders and verifiers),” says the blog. “Indicio DEGov enables the natural authority for a network or use case to orchestrate interaction by publishing a machine-readable governance file.”

The orchestrated files can be configured to “respect hierarchical levels of authority,” enabling “seamless interaction driven by automated trust.”

The vision is to create a way to enable massive networks of automated algorithmic entities to be able to interact with the proper permissions.

“Every element of this system can be known, can authenticate another element, and can share data in complex workflows,” Indicio says. “Each interaction can be made secure and element-to-element.”

Article Topics

agentic AI orchestration  |  AI agents  |  data protection  |  GDPR  |  Indicio  |  Nvidia  |  verifiable credentials