Global Fintech Fest 2026: Enforceable Privacy for Agent Collaboration

Financial agents need to collaborate across systems that hold sensitive information. Making those connections useful means giving institutions control over what can be computed and what each participant can learn.

At Global Fintech Fest 2026 in Mumbai, our Agent Privacy-Preserving Protocol, AP3, was selected among the top submissions for the Trustworthy Agentic AI thesis. Ambika Pande also joined the panel on AI, Innovation, and Inclusion.

A privacy layer for agent collaboration

The panel explored a question central to our work: how can agents collaborate across organisations and trust boundaries without exposing the underlying data?

Communication protocols help agents exchange messages. A financial workflow also needs a way to protect the inputs used to reach a decision. AP3 addresses that requirement through computation over encrypted inputs, allowing participants to work together while keeping their private data protected.

We design AP3 to work alongside the wider agent ecosystem. Teams should be able to add private computation to a workflow without having to replace the tools they use for agent communication.

From a protocol to an enterprise workflow

GFF week also created opportunities to discuss collaboration requirements with enterprises. Those requirements help shape useful implementation questions: which parties contribute data, what result is needed, and how should the workflow enforce the agreed boundaries?

We continued the conversation at the LF Decentralized Trust India community event during the same week, where Ambika presented AP3. Together, these engagements connected privacy architecture with the people building financial services around agents.

Silence Laboratories recap · Festival dates