For the past two years, the conversation around enterprise AI has predominantly focused on the 'model war' – a race between giants like OpenAI's GPT, Anthropic's Claude, and Google's Gemini to develop the most powerful foundational models. However, new insights suggest that the next major strategic battle for enterprise adoption is moving beyond the raw intelligence of these models and into the operational machinery that controls them: the agent control plane.
What Happened
According to new VB Pulse survey data, the enterprise AI landscape is witnessing a crucial shift. The survey, which tracks the preferences of technical decision-makers, indicates that the competition is intensifying around agentic orchestration platforms – the layer where AI agents plan, call tools, access data, manage workflows, and ensure compliance with security protocols.
Leading the charge in this critical space are Microsoft's Copilot Studio and Azure AI Studio, which captured 38.6% primary-platform adoption in February, a notable increase from 35.7% in January. OpenAI's Assistants and Responses API followed, securing second place with 25.7%, up from 23.2%.
Significantly, Anthropic, known for its Claude models, made its first measurable appearance in this tracker. Moving from 0% in January to 5.7% in February for Anthropic tool use and workflows, this marks a strategic entry into native orchestration beyond merely offering foundational models.
Image 1: Control room for AI neural network diagram: image omitted due to site embedding policy; open the original article (VentureBeat) (opens in a new tab) to view it. Photo/source: VentureBeat (opens in a new tab).
This early data, though based on a smaller cohort of 70 respondents, signals a clear trend: enterprises are not just evaluating chatbot capabilities; they are deciding on the infrastructure that will host their live AI operations. This decision dictates whether an AI workflow will reside within Microsoft's ecosystem, OpenAI's, or Anthropic's.
Image 2: VP Pulse Enterprise Agentic Orchestration change in respondents' primary agent orchestration platform from Jan-Feb 2026: image omitted due to site embedding policy; open the original article (VentureBeat) (opens in a new tab) to view it. Photo/source: VentureBeat (opens in a new tab).
Why It Matters
For developers, IT architects, and enterprise decision-makers, this shift is profound. The choice of an agent orchestration platform determines far more than just the AI model's performance on a benchmark. It dictates:
- Operational Control: Where and how AI agents are deployed, monitored, and managed in production environments.
- Tooling and Integration: The ease with which agents can connect to internal enterprise systems, databases, and external APIs to perform complex tasks.
- Data Access and Security: How agents securely access and process sensitive enterprise data, ensuring compliance and preventing unauthorized actions.
- Workflow Automation: The capability to design, implement, and scale multi-step AI-driven workflows that integrate seamlessly into existing business processes.
- Governance and Auditing: The ability to prove to security and compliance teams that agents operate within defined guardrails and do not execute unintended actions.
Essentially, choosing an orchestration layer is about selecting the underlying operating system for your enterprise AI. It's a strategic decision that impacts architectural flexibility, vendor lock-in, long-term scalability, and the overall reliability and safety of AI applications within an organization. For companies building serious AI applications, the model is just one component; the orchestration layer is where the real work happens.
What To Watch
As the enterprise AI battle heats up in the agent control plane, several key areas will be crucial to monitor:
- Anthropic's Expansion: How quickly can Anthropic build out and promote its native orchestration capabilities to rival the established offerings from Microsoft and OpenAI? Their initial foothold is small but significant.
- Google's Strategy: Will Google, with its Gemini models, make a more forceful entry into the agent orchestration space with dedicated enterprise platforms, or will it primarily focus on model improvements?
- Open-Source Alternatives: As this becomes a critical layer, will we see more robust open-source frameworks and platforms emerge for agent orchestration, offering enterprises greater flexibility and control?
- Security and Compliance Features: The importance of robust security, auditing, and governance features within these control planes cannot be overstated. Providers that can offer the most trusted and transparent environments will likely gain a competitive edge.
- Developer Experience: The platform that offers the most intuitive and powerful tools for developers to build, test, and deploy agents will be key to winning mindshare and adoption.
The era of purely debating AI model superiority is giving way to a more holistic view where the entire operational stack, particularly the agent control plane, will determine who leads the enterprise AI revolution.