OSPOlogy + OSPO Summit

Open Invention Network has facilitated the patent non-aggression community for open source for over twenty years. In January 2026, it announced an evolution of its community called OIN 2.0. This evolution is designed to support long-term sustainability, greater coverage for emerging open source technologies, and new opportunities for support against threats like Non-Practicing Entities (NPEs). While IP departments have traditionally focused on patent matters, OSPOs are a critical bridge between open source knowledge, product teams and legal. Understanding how OIN 2.0 works and how it fits into broader open source governance issues is a key part of comprehensive open source management. This talk will empower OSPO team members to work fluently in this space.

Open Source is everywhere and soon "multiplied by AI". The continuous pressure from security attackers, digitalization of everything and global geopolitical uncertainties create a higher need for transparency and trustful business relations.
Since a decade the global OpenChain Project community is continuously working on specifications and standards with the goal to build trust in the supply chain.
As trust is a human need, this is not only paperwork but also consists of a "community of conformance" with public contacts from conformant organizations to whom you can address any detected Open Source supply chain concern as well as a variety of workgroup-meetings and events where people share good practices and collaborate on improvements – all joint in the spirit of Open Source and the will to fix the issues at the source.
In this talk Marcel Kurzmann (OpenChain Board Member on behalf of Bosch) and Mary Meixia Wang (ED of OpenChain) will give an overview of the opportunities for chinese organizations to adopt the OpenChain Standards, become collaborating part of the global OpenChain "family" and thus increase their fitness for the global supply chains.

A structural divergence is quietly reshaping the global open source landscape. On one side is what we call native open source — the global commons governed by community meritocracy, voluntary contribution, and fork-based exit rights. On the other side is local open source — state-sponsored foundations , government-certified AI agent registries , and enterprise consortia operating under administrative logic rather than community governance.

These two systems look alike on the surface (both produce code, both use open licenses), but their institutional logic is fundamentally different — and becoming increasingly incompatible.

For enterprises operating OSPOs in China — especially multinational corporations, state-owned enterprises, and Chinese-born global companies — this divergence poses a practical dilemma:

1. Compliance vs. Contribution:Which upstream communities count as "safe" for regulatory compliance? Can you contribute to both systems without triggering conflict?
2. Talent Pipeline
3. Supply Chain Risk
4. OSPO Role Evolution: Is the OSPO becoming a "translation layer" between two incompatible institutional logics? What new capabilities does that require?

Legacy. We all have to deal with it in one form or another, and it is never fun. Especially organizational legacy. And old companies have a lot of it. But just because it may be hard, does not mean that it is impossible.

This talk is about building an Open Source Program Office (OSPO) and how you can do it even if your organization does not have the privilege of starting on a green field. It focuses on what an OSPO is, why it gained popularity in recent years, and why it might be a good idea for you to establish one. The talk draws heavily on the experience of Bosch in creating an OSPO out of existing open source-related activities with the aim that it may be used as a blueprint by others.

As the open-source ecosystem scales, traditional manual or rule-based governance tools are hitting efficiency bottlenecks. The rise of large language models (LLMs) offers a new paradigm.

This session will unpack how Ant Group's Open Source Program Office (OSPO) has built and integrated an AI-driven infrastructure for daily project governance. We will move beyond the hype and explore the real-world architecture and engineering practices, covering:

• Intelligent Pipelines: Building LLM-powered bots for automated issue triaging, assignment, and PR pre-reviews.
• Risk Management: AI-assisted code provenance and license compliance scanning.
• Data Insights: Upgrading from traditional dashboards to an NLP-based "Community Health AI Diagnostic System."

OSPOs must evolve at pace to enable the AI Landscape and this means both evolving practices within the OSPO itself and building better understanding of AI.
This session is an opening session, offering headline updates on the global AI openness landscape, to set the scene for the day's conversations.
Using OpenUK and OpenHQ's deep research and reporting on global AI Openness Amanda will update the audience on the key challenges faced by open source from AI and on progress in AI Openness in 2026.
The session will offer both an essential update and practical, contextualised insights critical to the audience.

Many companies face practical challenges when establishing an OSPO (Open Source Project Management Office), such as limited manpower, insufficient understanding from business departments, and high communication costs, making it difficult to quickly transition to routine operations.
This topic will share practical experience with internal open source portals and AI Agent applications, demonstrating how to leverage AI technology to quickly implement a three-step strategy for OSPO deployment: establishing communication and collaboration channels, identifying entry and exit points, and using open source as a medium to support ecosystem development.

Constraints such as limited hardware, finite context capacity, and the cost of complex agent workflows are pushing teams beyond isolated model interactions. Around them, a broader layer of open infrastructure is emerging: agent harnesses, MCP, reusable skills, gateways, shared context, and mechanisms for coordinating actions across tools

This session explores whether AI progress may come not only from ever-larger frontier models, but also from more distributed forms of intelligence, where specialized agents, open or local models, tools, and data sources can exchange context and collaborate. Using Goose, MCP, AGENTS.md, and agentgateway, we will examine what this stack could mean for OSPOs evaluating dependencies, interoperability, governance, and upstream participation

Join us for drinks and appetizers with your fellow speakers and attendees. All attendees of OSPOlogy + OSPO Summit are welcome.