institutional-leaders/anthropic-krieger-mcp-infrastructure-critical

Quote / Position

“MCP started as an internal project to solve a problem our own teams were facing. When we open sourced it in November 2024, we hoped other developers would find it as useful as we did. A year later, it’s become the industry standard for connecting AI systems to data and tools, used by developers building with the most popular agentic coding tools and enterprises deploying on AWS, Google Cloud, and Azure. Donating MCP to the Linux Foundation as part of the AAIF ensures it stays open, neutral, and community-driven as it becomes critical infrastructure for AI. We remain committed to supporting and advancing MCP, and with the Linux Foundation’s decades of experience stewarding the projects that power the internet, this is just the beginning.”

Context

Krieger’s framing is the load-bearing vendor-side statement of the open-by-infrastructure current. Five structurally important moves:

  1. Internal-tool → industry standard. MCP started inside Anthropic; one year later 10,000+ public MCP servers, 97M monthly SDK downloads, adopted by ChatGPT / Cursor / Microsoft Copilot / Gemini / VS Code. The internal-tool-to-industry-standard arc is the field guide’s parallel to JoinData’s cooperative-to-standards-body path.
  2. “Critical infrastructure for AI.” MCP is now classified as critical infrastructure, not a vendor product. The donation to Linux Foundation’s AAIF is the institutional move that makes it so — same institutional pattern as Kubernetes, PyTorch, Linux Kernel itself.
  3. Platinum member of AAIF (alongside AWS, Block, Bloomberg, Cloudflare, Google, Microsoft, OpenAI). The vendor-side institutional commitment is at the same tier as the open-by-state-funding programmes in the EU and India.
  4. Vendor neutrality as substantive claim. “Stays open, neutral, and community-driven.” Mozilla’s “the harness is the new frontier” finding + Anthropic’s “Linux Foundation neutral stewardship” claim + AAIF’s platinum-tier roster together establish that the harness layer is the institutional contest.
  5. Python + TypeScript SDK adoption (97M monthly downloads) — the developer-ecosystem metric that Mozilla uses for “79% use, 51% deployed” with broader weight — shows how the harness layer is achieving deployment parity where the model layer still shows the gap.

This entry complements Mozilla-krikorian-open-source-turning-point.md (Mozilla CTO industry-side frame), the osi-maffulli-binary-threshold-open-source-ai.md (definitional side), and ec-virkkunen-european-sovereign-ai-vision.md (EU institutional side).

Application to agrifood

The agrifood translation: Mozilla’s “the harness is the new frontier” finding lands hardest in agrifood, where data-rights and permission frameworks have been ahead of the broader AI debate. An MCP-style agentic harness for agrifood — covering OADA / JoinData / NAPDC / OCAP / cooperative schemas as exposed tools — is the agrifood counterpart to “MCP for AI software engineering.” Gap G-364 (AAIF-equivalent for agrifood) and G-365 (agrifood-specific permission writing-up) track this absence.

Source

Mike Krieger, Chief Product Officer, Anthropic. Quoted in Linux Foundation press release Linux Foundation Announces the Formation of the Agentic AI Foundation (AAIF), Anchored by New Project Contributions Including Model Context Protocol (MCP), goose and AGENTS.md. December 9, 2025. https://www.linuxfoundation.org/press/linux-foundation-announces-the-formation-of-the-agentic-ai-foundation