Provision record
Salesforce Einstein · Salesforce Trusted AI Principles · View original document ↗

Data Masking of PII Before LLM Transmission

Medium severity Medium confidence Explicit document language Unique · 0 of 352 platforms
Stay ahead of the changes
Track Salesforce Einstein and get the diff the day its terms change.
Share 𝕏 Share in Share 🔒 PDF
Document Record

What it is

This provision describes a technical process that replaces PII and proprietary business data with non-identifiable tokens before prompts are transmitted to the LLM, with the original data restored after the response is generated.

ⓘ

This analysis describes what Salesforce Einstein's agreement states, permits, or reserves. It does not constitute a legal determination about enforceability. Regulatory applicability and practical outcomes may vary by jurisdiction, enforcement context, and individual circumstances. Read our methodology

ConductAtlas Analysis

Why it matters (compliance & governance perspective)

This provision describes a technical de-identification mechanism that operates on data in transit to third-party LLMs, which is directly relevant to compliance obligations under GDPR, CCPA, and HIPAA regarding the processing of personal and sensitive data by AI systems. The effectiveness of this mechanism as a de-identification or anonymization control under applicable law depends on the specific tokenization methodology and whether re-identification risk is adequately mitigated.

⚠

Interpretive note: The adequacy of the described tokenization mechanism as de-identification or anonymization under GDPR, CCPA, and HIPAA depends on technical implementation details not fully specified in the document.

Consumer impact (what this means for users)

Under this provision, PII and proprietary business data in prompts are replaced with non-identifiable tokens before transmission to third-party LLMs, with the document stating this process shields confidential information while preserving response quality. Enterprise customers processing sensitive personal data through Salesforce AI features should assess whether this tokenization mechanism satisfies their specific de-identification or anonymization obligations under applicable data protection law.

Cross-platform context

See how other platforms handle Data Masking of PII Before LLM Transmission and similar clauses.

Compare across platforms →
▸ View Original Clause Language DOCUMENT RECORD
"
Complementing this, data masking is the process that replaces sensitive Personally Identifiable Information (PII) or proprietary business data with non-identifiable tokens before the prompt is sent to the LLM. This shields confidential information while still providing the necessary context for the LLM to generate a personalized and useful response.

Excerpt from Salesforce Einstein's Salesforce Trusted AI Principles

ConductAtlas Analysis

Institutional analysis (regulatory & governance intelligence)

(1) REGULATORY LANDSCAPE: This provision engages GDPR pseudonymization and anonymization standards, CCPA definitions of personal information and de-identified data, and HIPAA Safe Harbor and Expert Determination de-identification standards for covered entities.

Insight

Unlock the full institutional analysis

Enforcement risk, jurisdiction flags, contract triggers, and due diligence action items.

Applicable agencies

  • Federal Trade Commission (ftc)
    Oversees unfair or deceptive business practices and can investigate companies that mislead consumers about data collection, sharing, or use.
    Who can file: Anyone affected by the company's practices (US or international)
    What you need: Your account details, a timeline of relevant events, and a description of the specific issue
    What to expect: Complaints inform FTC enforcement priorities and investigations but do not result in individual resolution or compensation
    File a complaint →
  • Department Of Health & Human Services, Office For Civil Rights (hhs Ocr)
    Enforces HIPAA Privacy and Security Rules, which protect health information held by healthcare providers, health plans, and their business associates.
    Who can file: Anyone whose HIPAA rights may have been violated by a covered entity (healthcare provider, health plan, or healthcare clearinghouse)
    What you need: Name of the entity, description of the violation, date of the incident, and your contact information. Must file within 180 days of the violation.
    What to expect: HHS OCR investigates and may require the entity to take corrective action. Does not provide individual compensation. Serious violations can result in civil monetary penalties.
    File a complaint →

Provision details

Document information
Document
Salesforce Trusted AI Principles
Entity
Salesforce Einstein
Document last updated
May 12, 2026
Tracking information
First tracked
July 12, 2026
Last verified
July 12, 2026
Record ID
CA-P-074447
Document ID
CA-D-00818
Evidence Provenance
Source URL
Wayback Machine
Content hash (SHA-256)
c9bb51f7a29871aea2e399cd45ac7de48db93c4bdcfc153b82b72422b3556af6
Analysis generated
July 12, 2026 16:47 UTC
Methodology
Evidence
✓ Snapshot stored   ✓ Hash verified
Citation Record
Entity: Salesforce Einstein
Document: Salesforce Trusted AI Principles
Record ID: CA-P-074447
Captured: 2026-07-12 16:47:16 UTC
SHA-256: c9bb51f7a29871ae…
URL: https://conductatlas.com/platform/salesforce-einstein/salesforce-trusted-ai-principles/provision/CA-P-074447/data-masking-of-pii-before-llm-transmission/
Accessed: Sept. 26, 2026
Permanent archival reference. Stable identifier suitable for legal filings, compliance documentation, and research citation.
Classification
Severity
Medium
Categories

Other risks in this policy

Get the research letter

Companies change their terms quietly. We read every version and catch what actually changed. One email a week on the changes that matter and what they mean.

Frequently Asked Questions

What does Salesforce Einstein's Data Masking of PII Before LLM Transmission clause do?

This provision describes a technical de-identification mechanism that operates on data in transit to third-party LLMs, which is directly relevant to compliance obligations under GDPR, CCPA, and HIPAA regarding the processing of personal and sensitive data by AI systems. The effectiveness of this mechanism as a de-identification or anonymization control under applicable law depends on the specific tokenization methodology and …

How does this clause affect you?

Under this provision, PII and proprietary business data in prompts are replaced with non-identifiable tokens before transmission to third-party LLMs, with the document stating this process shields confidential information while preserving response quality. Enterprise customers processing sensitive personal data through Salesforce AI features should assess whether this tokenization mechanism satisfies their specific de-identification or anonymization obligations under applicable data protection …

Is ConductAtlas affiliated with Salesforce Einstein?

No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by Salesforce Einstein.