## Salesforce Tackles the AI 'Wild West' with New Open-Source Language

Salesforce just made a bold move to standardize the chaotic world of enterprise AI development. At its TDX 2026 conference, the company announced the open-sourcing of Agent Script, a new TypeScript-based framework for building and orchestrating AI agents. The goal: to move enterprises from ad-hoc, experimental bots to predictable, production-grade AI systems.

Agent Script provides a schema-driven language that allows developers to define an agent's behavior, mixing deterministic rules with the more unpredictable, creative power of large language models. This hybrid approach is designed to give developers the control they need to ensure agents are reliable and safe, while still allowing for the flexibility that makes AI so powerful.

Alongside Agent Script, Salesforce also launched Agentforce Labs for rapid prototyping and a new set of Agent Development Life Cycle (ADLC) skills. The ADLC framework, which Salesforce has been developing internally, emphasizes observability and repeatability, two critical components for building enterprise-ready AI. The message from Salesforce is clear: the era of throwing a model at a problem and hoping for the best is over.

For CISOs, this is a welcome development. The rapid, often unsanctioned, adoption of AI tools has created a new world of "shadow AI," with little visibility or control. A standardized framework like Agent Script, coupled with a focus on the full development lifecycle, could provide the guardrails needed to innovate safely. While the real-world impact remains to be seen, Salesforce's move is a significant step toward bringing much-needed discipline to the enterprise AI landscape.