The 2026 enterprise agentic AI orchestration-framework choice across the five major frameworks (AWS Bedrock AgentCore GA October 2025, Microsoft Azure AI Foundry plus Copilot Studio, Google Vertex AI Agent Builder plus the open-source Agent Development Kit, OpenAI Agent Builder GA October 2025 plus the open-source Swarm primitive, Anthropic Claude Agent SDK late 2024 plus the open-source Model Context Protocol) prices the decision as a 3-year orchestration-layer commitment along five comparison axes (orchestration primitive, tool-use protocol, deployment topology, observability tier, exit cost), with the framework choice resolved by gravity-fit against the customer's existing cloud, identity, and data substrate rather than by model-tier performance; the 2026 cross-vendor convergence on the Model Context Protocol as the tool-use standard is the structural change that makes exit cost newly tractable for customers who write tool definitions in the protocol and emit OpenTelemetry traces, materially lowering the year-three re-platforming cost relative to the framework-native alternatives.

Anchored on (a) AWS Bedrock AgentCore product documentation (GA October 2025 with Runtime, Memory, Identity, Gateway, Code Interpreter, Browser, Observability primitives); (b) Microsoft Azure AI Foundry product documentation (late 2024 rebrand from Azure AI Studio; Copilot Studio multi-agent orchestration 2025 expansion; November 2024 autonomous agents launch; 2025 Microsoft Agent 365 management surface); (c) Google Vertex AI Agent Builder + ADK (open-sourced 2024) + Agents Garden managed catalogue; (d) OpenAI Agent Builder GA October 2025 + Assistants API persistence layer + open-source Swarm primitive; (e) Anthropic Claude Agent SDK late 2024 + Model Context Protocol November 2024 launch (modelcontextprotocol.io) + 2025 cross-vendor adoption. The competitive-position characterisation is from current vendor documentation; future framework reframes (a Foundry-to-something-else rename, an AgentCore architecture pivot, the Vertex AI roadmap acceleration) would change the matrix. 60-day review cadence (26 Jul 2026). Trigger conditions: (1) major vendor announcing structural framework reframe moves toward Partial; (2) Model Context Protocol or alternative tool-use standard achieving genuine cross-vendor portability (not just adoption) changes exit-cost axis materially; (3) published independent (non-vendor-funded) benchmark comparing the five frameworks on enterprise deployment outcomes (latency, cost, error rates at production scope) hardens or weakens competitive position claims; (4) major partnership shifts (Microsoft-OpenAI restructure, AWS-Anthropic deepening, Google first-party competitiveness narrowing) reshape comparison. Sibling pairwise /compare/ pages cover model-tier comparisons (orchestration layer constant, model variable); AM-169 covers the protocol-tax piece (MCP vs A2A vs Llama Stack) at the protocol-level decision.