Which component of the Einstein Trust Layer ensures alignment of AI prompts with Salesforce records?

The component of the Einstein Trust Layer that ensures alignment of AI prompts with Salesforce records is known as grounding in CRM data. This mechanism allows AI models to link their responses and suggestions directly to the data within Salesforce, enhancing their relevance and accuracy. By leveraging existing Salesforce records, this grounding capability provides contextually appropriate responses that are informed by real-time data, facilitating a more informed and personalized interaction for the user.

Grounding in CRM data is vital for ensuring that the AI's outputs reflect the most current information available in the Salesforce environment, thereby helping users to make better decisions based on the insights derived from their organizational data. This connection enhances the overall trustworthiness of AI interactions, as users can rely on the information presented as being directly sourced from Salesforce records.

In contrast, the other options, while potentially useful in various contexts within AI frameworks, do not specifically address the requirement for alignment with Salesforce records. Dynamic grounding might refer to adjusting AI responses based on changing inputs, data masking deals with obfuscating sensitive information, and instruction defense focuses on protecting against misleading prompts or requests. However, none of these options provide the direct integration and contextual grounding to Salesforce data that is central to effective AI interaction within the platform.