Salesforce acquired DataRobot’s synthetic data division to bolster Einstein AI and Data Cloud with privacy-compliant synthetic CRM datasets. The practical bet: faster testing and model training without putting sensitive customer records directly in the development loop.
Salesforce acquires DataRobot’s synthetic data division for privacy-safe AI development
Salesforce announced it has acquired DataRobot’s synthetic data division, positioning the assets as an accelerator for its Einstein AI platform and Salesforce Data Cloud. The stated focus is enabling customers to generate realistic synthetic CRM data for testing and AI training while reducing exposure of personal or sensitive customer information.
The move lands as enterprises face tighter expectations around data handling and documentation. Salesforce is framing synthetic data as a privacy-first mechanism that can support development workflows—especially where teams would otherwise rely on production-like customer records for QA, sandbox environments, and model iteration.
- Faster iteration with less risky data access: If synthetic datasets are “good enough” for common testing and training tasks, teams can shorten approval cycles that typically slow down access to real customer data.
- Privacy and compliance posture improves on paper—if governance is real: Synthetic data can reduce direct exposure to personal data, but teams still need controls, documentation, and validation to show the synthetic outputs are privacy-compliant.
- Platform lock-in becomes more likely: Bundling synthetic data generation into Data Cloud/Einstien AI could pull more of the data lifecycle into Salesforce, affecting tool choice for data engineering, MLOps, and governance teams.
- New evaluation burden for data leaders: Expect internal questions about fidelity vs. privacy tradeoffs, acceptable use cases (testing vs. training), and how synthetic data fits with existing policies for GDPR, CCPA, and HIPAA-aligned programs.