Salt Security announced Salt Code, a product designed to apply security and compliance policies to code generated through enterprise AI coding assistants such as Claude, Cursor, GitHub Copilot, Windsurf, Codex, and Gemini CLI.
Salt Security announced Salt Code, a new component of its Agentic Security Platform intended to enforce security and compliance policies inside enterprise AI coding assistant workflows.
In a press release distributed by PRNewswire and published on Salt Security’s website, the company described Salt Code as an “agentic security solution” for organizations using AI coding assistants. Salt said the product is designed to work across tools including Claude, Cursor, GitHub Copilot, Windsurf, Codex, and Gemini CLI.
Salt’s stated aim is to help enterprises apply security rules earlier in the software development process, before AI-generated code is merged or deployed. The company said Salt Code enforces policies across the AI-generated-code development lifecycle, rather than relying only on later review or runtime controls.
The announcement positions the product as part of Salt Security’s broader Agentic Security Platform. Salt’s product page says organizations can connect Salt Code to AI coding assistants via MCP, referring to the Model Context Protocol, and use prebuilt secure coding packs.
According to Salt Security’s Salt Code product page, the product is “free to start” and includes four secure coding packs. Salt says those packs contain more than 40 policies intended to guide AI coding assistants toward policy-compliant generated code.
The company describes the product as a way to provide AI coding assistants with security and compliance context at the time code is being generated. Salt’s announcement says this approach is meant to reduce the risk that AI-assisted development produces code that violates organizational rules, introduces insecure patterns, or fails to meet compliance expectations.
Salt did not provide independent benchmark results in the cited announcement or product page. The available sources are company materials and a PRNewswire release based on Salt’s announcement, so claims about effectiveness should be read as Salt Security’s own positioning rather than third-party validation.
The release reflects a growing enterprise concern: AI coding tools can accelerate development, but they also create new governance questions. Organizations adopting assistants such as GitHub Copilot, Cursor, Claude, Gemini CLI, Codex, and Windsurf need ways to manage what those tools generate, what rules they follow, and how security guidance is applied before code reaches production systems.
Salt’s approach focuses on policy enforcement inside the assistant workflow itself. If implemented as described by the company, that could shift some secure-coding controls from post-generation scanning into the prompt-and-generation stage of software development. That model may appeal to security teams trying to standardize how developers use multiple AI coding assistants across an enterprise.
Salt’s announcement does not include public third-party testing, customer case studies, pricing beyond the “free to start” language, or detailed technical documentation in the cited materials. It also does not specify how the product handles conflicts between assistant behavior, developer instructions, and policy requirements across each supported coding tool.
For now, the launch is best understood as Salt Security’s entry into a fast-developing category: security controls for AI-assisted software development. The company is betting that enterprises will want policy enforcement closer to where AI-generated code is produced, not only in later scanning, review, or deployment stages.
Salt Security announced Salt Code, a new component of its Agentic Security Platform intended to enforce security and compliance policies inside enterprise AI coding assistant workflows.
What Salt announced In a press release distributed by PRNewswire and published on Salt Security’s website, the company described Salt Code as an “agentic security solution” for organizations using AI coding assistants.
Salt said the product is designed to work across tools including Claude, Cursor, GitHub Copilot, Windsurf, Codex, and Gemini CLI.
Continue reading
