Two major industry bodies, Trust Over IP and the Decentralized Identity Foundation, have joined forces to address one of AI’s biggest challenges – trust. They have launched three new Working Groups aimed at building standards and frameworks that ensure AI agents can operate reliably and responsibly and be depended on by third parties.
The new groups are:
- The Joint ToIP/DIF Decentralized Trust Graph Working Group (DTGWG)
- The ToIP AI and Human Trust Working Group (AIMWG)
- The DIF Trusted AI Agents Working Group (TAIAWG)
The initiative reflects the growing urgency of bringing AI into the fold of many functions in society, whilst managing the risks associated with privacy.
“As AI accelerates, the question is no longer if we need digital trust frameworks, but how quickly we can deliver them,” quoted LFDT Executive Director Daniela Barbosa.
The Decentralised Trust Graph Working Group is aiming to tackle the core purpose of identification and confirm a real human behind the identity without compromising privacy. A blog post written by Drummond Reed, Wenjing Chu and Andor Kesselman shares that the group aims to mitigate centralised databases and instead allow a privacy-first approach to individuals managing their own verifiable data transitions through AI agents and digital wallets. By building a “decentralised trust graph of verifiable trust relationships”, it allows all parties to control their “own portable subgraph of trust relationships in their own digital agents and wallets”.
The AIMWG, evolving from a task force launched in 2022, is focused on embedding trust into AI itself. The group is examining how ToIP’s Trust Spanning Protocol can enhance interactions between humans and AI agents, while also documenting use cases and frameworks for delegation, accountability, and identity.
The TAIAWG is addressing the technical foundations required for autonomous agents to act responsibly on behalf of individuals and organisations. The group will develop specifications and governance models to ensure AI agents can carry out tasks such as data exchange, decision-making, and transactions within a secure, zero-trust framework. Its first deliverable will be a set of Agentic Authority Use Cases, focusing on how authority can be delegated safely and efficiently across agent workflows.
