Salesforce's best-in-class security guardrails are natively embedded in Einstein. Learn how the Einstein Trust Layer addresses the data privacy challenges of generative AI.
Sessions
Explore Dreamforce 2023 sessions.
View 1,500+ sessions by role, topic, or industry to see an example of what's to come at Dreamforce 2024.
Dig Deep into Data Privacy with the Einstein Trust Layer
, Associate Product Marketing Manager, Salesforce
, Product Management Senior Director, Salesforce
Role: IT
Product: Einstein
- Tuesday, Sep 122:30 PM - 2:50 PM PDT(9:30 PM - 9:50 PM GMT)Einstein Theater (Moscone West, Level 2)
- Wednesday, Sep 1311:30 AM - 11:50 AM PDT(6:30 PM - 6:50 PM GMT)Olympic Theater (Moscone West, Level 2)
Session Summary by Einstein: In this Dreamforce session, Salesforce product marketer Kamaria Wilson and product manager Gene Becker discussed the importance of data privacy and generative AI, focusing on the Einstein trust layer. They highlighted the AI trust gap, where companies are apprehensive about the technology due to potential inaccuracies, lack of context, data security concerns, risk of toxicity and bias, and integration issues. To address these concerns, Salesforce introduced the Einstein trust layer, a platform for deploying trusted and secure AI. This includes secure data retrieval, dynamic grounding, data masking, toxicity detection, auditing, and zero data retention. They also discussed the importance of prompt templates, data privacy and security, and rigorous evaluation for ethical use. The session emphasized that trust is built into the fabric of all Salesforce products, ensuring that generative AI can be deployed with confidence.
Key Takeaways by Einstein:
- • Salesforce is addressing the AI trust gap with the Einstein trust layer, a platform for deploying trusted and secure AI, which includes features like secure data retrieval, dynamic grounding, data masking, toxicity detection, and zero data retention.
- • Salesforce's generative AI platform is built on Hyperforce, ensuring built-in security, enterprise scale, and regulatory compliance, and includes a secure gateway for communication with third-party model partners like OpenAI.
- • Salesforce's trust layer technologies provide structure and security to improve the accuracy of large language models, including prompt templates, system prompts, data hydration, data masking, toxicity detection, feedback and audit trail, and rigorous evaluation for ethical use.
From thought-provoking sessions to hands-on workshops, Dreamforce equips you with the tools and knowledge to succeed."
Nadina D. Lisbon
CRM Enterprise Architect
NetApp
