OPAQUE Announces Applications for AgenTrust Fellowship to Advance Open Standards for Verifiable AI Agents
OPAQUE Announces Applications for AgenTrust Fellowship to Advance Open Standards for Verifiable AI Agents
Six-month paid fellowship invites select security engineers, researchers, and compliance practitioners to help build the open standards for trustworthy AI agents safely
SAN FRANCISCO — July 28, 2026 — OPAQUE today announced that applications are opening on August 1 for the inaugural AgenTrust Fellowship, a six-month paid, part-time (20 hrs/week) program for engineers, researchers, and governance practitioners to help build the open infrastructure behind verifiable AI. The first cohort will include three to five fellows and begin in September 2026, and it does not require selected participants to quit their jobs. Selected participants will work alongside OPAQUE Chief Platform Officer and Agent Governance Toolkit (AGT) creator Imran Siddique, contributing directly to the open AgenTrust ecosystem and helping advance the next generation of verifiable AI standards.
As AI agents become part of enterprise infrastructure, the industry's focus is expanding beyond model capability to building AI that organizations can confidently deploy, govern, and scale. That shift requires open standards that work across vendors, clouds, and ecosystems rather than proprietary approaches. AgenTrust was created to help build that foundation through open, independently verifiable standards rather than vendor-defined claims.
"Today, enterprises have no standard, independent way to prove that an AI agent actually followed policy once it has been deployed," said Imran Siddique, Chief Platform Officer at OPAQUE. "That isn't a product problem, it's a standards problem. Solving it requires an open community to build common ways to verify identity, enforce policies, and provide runtime evidence across the industry. The AgenTrust Fellowship exists to grow that community and accelerate the standards AI will ultimately depend on."
The fellowship builds on the momentum behind Siddique’s AGT, an open-source governance layer for AI agents with almost 5,000 GitHub stars, nearly 1,000 conformance tests, and integrations across major agent frameworks. AgenTrust extends that work into an open standards ecosystem where fellows will help solve problems such as proving that a model's weights never leave a secure enclave or building policy-enforced inference hosting for regulated industries.
Over six months, fellows will contribute across the AgenTrust open-source ecosystem, tackling real-world challenges in AI governance, confidential computing, and runtime verification. Their work may include advancing open standards for agent identity, protecting sensitive models and data during execution, and developing cryptographic evidence that allows enterprises and regulators to independently verify how AI systems behaved across open AgenTrust projects such as TRACE, Agent Manifest, Confidential MCP, and the AGT.
Who Should Apply: The fellowship is intended for security engineers, distributed systems researchers, AI governance and compliance professionals, and Ph.D. students interested in shaping the future of open standards for trustworthy AI.
What Fellows Receive: Fellows will receive direct mentorship from Imran Siddique while contributing code, specifications, and technical guidance across the AgenTrust open-source ecosystem. Each fellow will complete a public technical artifact, such as a research paper, conference presentation, or engineering blog.
Applications Close August 29: The inaugural AgenTrust Fellowship is accepting applications starting on August 1; the full listing is here. The first cohort begins in September 2026. Applicants will be selected based on a technical proposal demonstrating how they plan to contribute to the AgenTrust ecosystem.
The fellowship extends OPAQUE's broader commitment to building AI governance in the open. By growing a community of contributors around the AgenTrust ecosystem, OPAQUE aims to accelerate the development of interoperable standards that enterprises, regulators, and developers can build on together. It also builds on OPAQUE's recent launch of AgenTrust open standards, including Agent Manifest, TRACE, Confidential MCP and Confidential A2A, and follows the company's expanded work with the Linux Foundation's Appia Foundation and Agentic AI Foundation, as well as contributions to the Coalition for Secure AI (CoSAI), to advance open standards for verifiable AI. Together, these efforts aim to build an open ecosystem where AI governance can be independently verified rather than vendor-defined.
About OPAQUE
OPAQUE is the Confidential AI company. Born from UC Berkeley's RISELab (now the Sky Compute Lab), OPAQUE lets organizations run AI models, agents, and workflows on their most sensitive data with hardware-rooted isolation and verifiable evidence that approved governance policies were actually enforced. Founded by Dr. Ion Stoica (co-founder of Databricks; co-director, UC Berkeley Sky Compute Lab), Dr. Raluca Ada Popa (ACM Grace Hopper Award winner; Senior Staff Research Scientist at Google DeepMind, where she leads AGI security research), and Rishabh Poddar (CTO); Imran Siddique, creator of the open-source Agent Governance Toolkit (AGT), is Chief Platform Officer. OPAQUE created the Confidential Computing Summit, now co-hosted with the Linux Foundation.
Media contact: opaquesystems@inkhouse.com and press@opaque.co