These tools and metrics are designed to help AI actors develop and use trustworthy AI systems and applications that respect human rights and are fair, transparent, explainable, robust, secure and safe.
Einstein Trust Layer
The Einstein Trust Layer is a secure AI architecture built into the Salesforce platform that provides guardrails to protect data privacy and security, improve the safety and accuracy of AI outputs, and enable Salesforce customers to use generative AI responsibly within Salesforce applications.
The Einstein Trust Layer is designed to allow organisations using Salesforce to benefit from generative AI while maintaining strict privacy and security controls over their company and customer data. It acts as an intermediary layer between user prompts and large language models (LLMs), ensuring that sensitive information remains protected when AI services are used within the Salesforce environment.
Within the Salesforce ecosystem, the Trust Layer introduces a set of technical mechanisms and policies that govern how prompts and responses interact with AI models. These mechanisms ensure that generative AI outputs are generated safely and in compliance with existing data protection and governance standards.
One key component of the Trust Layer is secure data retrieval and dynamic grounding, which allows prompts to be enriched with relevant business or CRM data stored in Salesforce while maintaining user permissions and access controls. This process helps ensure that AI responses are contextually accurate without exposing sensitive information.
The Trust Layer also incorporates data masking and prompt defense mechanisms that detect and mask sensitive data before prompts are sent to external language models. This helps prevent confidential or personally identifiable information from being exposed during AI interactions. Another key feature is zero data retention, which ensures that prompts and responses sent to third-party large language models are not stored or used for model training. In addition, the Trust Layer integrates toxicity detection and monitoring mechanisms that evaluate prompts and AI-generated outputs for harmful or inappropriate content and help create audit trails for oversight and governance.
Through these combined privacy, security, and governance mechanisms, the Einstein Trust Layer enables organisations using Salesforce to integrate generative AI capabilities into their workflows while maintaining strong protections for enterprise and customer data.
About the tool
You can click on the links to see the associated tools
Developing organisation(s):
Tool type(s):
Objective(s):
Impacted stakeholders:
Purpose(s):
Lifecycle stage(s):
Type of approach:
Maturity:
Usage rights:
Target groups:
Target users:
Stakeholder group:
Geographical scope:
People involved:
Required skills:
Risk management stage(s):
Technology platforms:
Tags:
- transparency
- data ethics
- ai security
- privacy
- ai safety
Use Cases
Would you like to submit a use case for this tool?
If you have used this tool, we would love to know more about your experience.
Add use case
Partnership on AI
