Ai

Microsoft Semantic Kernel dotnet 1.75.0 strengthens .NET validation and connector security

Microsoft Semantic Kernel dotnet 1.75.0 strengthens .NET validation and connector security

Microsoft has released Semantic Kernel dotnet 1.75.0, a focused update that strengthens .NET-side validation, improves connector safety, and refreshes platform dependencies. The release is especially notable for hardening REST API URL validation, tightening step-type checks, improving SQL and Redis escaping behavior, and updating SQL Server vector search support to the latest syntax.

What Changed

The dotnet 1.75.0 release updates the .NET package version and includes a broader dependency refresh. Beyond maintenance work, the most important changes are in validation and query safety. Microsoft hardened AllowedBaseUrls validation in RestApiOperationRunner, added validation for step types, and fixed several escaping issues affecting Redis text search values, SQL OBJECT_ID usage, and dynamic SQL string literals.

The update also modernizes SQL Server vector search support by aligning with the latest VECTOR_SEARCH() syntax. In addition, Microsoft removed MEVD components from the SK-release.slnf release solution filter, which may simplify the .NET release surface for maintainers.

Although the release is centered on .NET, the wider changelog also includes Python package version bumps, a blocklist extension for InMemoryCollection filter attributes, SQL Server connector escaping improvements, and an updated ipykernel requirement range.

Why It Matters

This version matters because it improves the reliability and security posture of Semantic Kernel deployments built on .NET. Harder validation around external API base URLs can reduce the risk of misconfigured or unsafe integrations, while better escaping in SQL and Redis-related paths helps prevent malformed queries and edge-case failures in production systems.

The SQL Server vector search syntax update is also important for teams building retrieval and AI orchestration workflows on Microsoft data infrastructure. It helps keep Semantic Kernel aligned with upstream database capabilities, which can reduce compatibility friction as vector search features evolve.

Overall, dotnet 1.75.0 is less about flashy new end-user features and more about safer execution, cleaner compatibility, and steadier platform foundations for enterprise AI applications.

Official Source: https://github.com/microsoft/semantic-kernel/releases/tag/dotnet-1.75.0