Salesforce AI Vendor Risk Profile
Enterprise CRM leader that combines proprietary AI models (Einstein, CodeGen, xGen) with OpenAI integration for Einstein GPT, embedding AI across sales, service, and marketing clouds.
Risk overview
Risk score: 13/100
Risk tier: Low
Safety rating: 87/100 (higher is better)
Lower risk scores indicate lower assessed risk.
Last verified: Apr 19, 2026 Aging
Risk dimensions
| Dimension | Risk score |
|---|---|
| Data Handling | 14/100 |
| IP Exposure | 10/100 |
| Jurisdiction | 13/100 |
| Security | 18/100 |
| Regulatory Compliance | 10/100 |
| Transparency | 10/100 |
| Business Stability | 10/100 |
| Dependency Chain | 14/100 |
| Agent Governance | 60/100 |
Analyst summary
Rating: Recommended
Salesforce's Einstein Trust Layer is the clearest example of an enterprise vendor engineering zero-retention pass-through to third-party LLM providers. For customers already on Salesforce, Agentforce and Einstein AI slot into existing compliance perimeters (SOC 2, ISO 27001, FedRAMP High) without new data-handling exposure.
Bottom line: A clean extension of Salesforce's existing trust envelope for Salesforce-standardized customers.
Strengths
- Einstein Trust Layer enforces zero-data-retention contracts with LLM providers
- Full compliance stack: SOC 2 Type II, ISO 27001/27017/27018/27701, FedRAMP High on Government Cloud Plus
- HIPAA BAA available for Health Cloud and eligible services
- Strong governance: audit trails, data masking, toxicity detection built into the Trust Layer
- Customer retains full ownership of Customer Data and AI-generated outputs derived from it
Concerns
- Value is unlocked only for existing Salesforce customers; AI-only adoption is not practical
- Agentforce and Data Cloud pricing is opaque and can escalate quickly
- Einstein AI features rely on upstream LLM providers, with capability tied to those relationships
Best for
- Salesforce-standardized enterprises adding AI to existing CRM, service, and marketing workflows
- Regulated industries where Salesforce's compliance posture is already in scope
- Customer service and sales teams needing AI within existing data-governance boundaries
Avoid if
- You are not already on Salesforce (the AI alone is not a reason to adopt it)
- You need general-purpose AI outside of CRM and customer data workflows
Citations
- Data Handling — Hipaa Baa Available
Salesforce offers a HIPAA Business Associate Agreement for Health Cloud and other eligible services.
- Data Handling — Outputs Feed Model Improvement
Salesforce's zero-data-retention architecture means no customer prompts or outputs are stored by third-party LLM providers.
- Data Handling — Trains On User Data
The Einstein Trust Layer ensures customer data is not retained or used to train models by third-party LLM providers.
- Ip Profiles — User Owns Outputs
Customer retains ownership of all Customer Data, including outputs generated from Customer Data by Einstein AI features.
- Jurisdiction Profiles — Incorporation Country
Salesforce, Inc. is a Delaware corporation headquartered in San Francisco, California (SEC Form 10-K).
- Security Compliance — Fedramp Authorized
Salesforce Government Cloud Plus holds FedRAMP High authorization.
- Security Compliance — Iso 27001
Salesforce holds ISO/IEC 27001, 27017, 27018, and 27701 certifications.
- Security Compliance — Soc2 Type2
Salesforce Services maintain SOC 2 Type II attestations available via the Compliance portal.