BACK TO NEWS

MCP Is Gaining Ground, but Governance and Security Still Need Work

By Protegrity
Apr 3, 2026

Summary

5 min
  • MCP is gaining traction, but enterprise adoption still comes with operational friction:
    The AI Business article explains that while the Model Context Protocol is emerging as a key open standard for connecting AI models to tools and data sources, organizations are still working through issues such as context bloat, fragmented server environments, and the complexity of scaling deployments responsibly.

  • Protegrity POV: open standards need strong security and governance guardrails:
    Bharat Kotian notes that MCP can improve productivity and help bring systems together, but successful adoption requires meaningful implementation effort, protection for sensitive data, and clear policy oversight as enterprise use grows.

A recent AI Business article explores how the Model Context Protocol (MCP) continues to gain traction as an open standard for connecting AI models to tools and data sources, even as teams encounter growing pains around context bloat, synchronization, and operational control. The piece also includes perspective from Bharat Kotian of Protegrity, who notes that MCP can improve productivity and bring systems together, but requires a meaningful time investment and a strong focus on protecting sensitive data.

Key takeaways

  • MCP is still maturing. Adoption has grown quickly across major cloud providers, but organizations are still working through issues like excessive context, fragmented server deployments, and the operational overhead of scaling MCP responsibly.
  • Security and policy oversight are essential. As Bharat Kotian points out, organizations experimenting with MCP need to ensure they maintain security around sensitive data, not just functionality and speed.
  • Control planes matter. The article highlights Uber’s use of an MCP gateway and registry to centralize control, showing how governance layers can reduce chaos across multiple MCP servers.
  • Ethics remains a gap. Even as technical adoption increases, the article notes that responsible AI governance still needs more attention at the protocol and deployment level.

What this means for enterprises

The takeaway is not that MCP is failing. It is that open standards need enterprise guardrails. As more teams connect agents, tools, and data through MCP, success will depend on having the right controls for identity, registry management, security policy enforcement, and ethical AI oversight.

Protegrity perspective. Open protocols like MCP can improve productivity and interoperability, but secure adoption depends on protecting sensitive data, maintaining policy oversight, and giving teams clear control points as deployments scale.

Note. This summary is based on the external article “MCP is Alive, but Faces Challenges” and is provided for convenience. Please refer to the original publication for full context and reporting details.