LangChain updated its Terms of Service to introduce a new deployment option called 'Bring Your Own Cloud Deployment' (BYOC), which allows customers to run the LangSmith Platform within their own cloud infrastructure while LangChain manages it through APIs without direct access to customer data. The company also clarified its Hybrid Deployment definition to specify that customer workloads execute in the customer's environment with traces and run data transmitted to LangChain's cloud, and expanded its commitment not to use customer data for training to explicitly include large language models. Additionally, the warranty disclaimer was revised to state that LangChain does not warrant the platform will be 'accurate' or 'complete' in addition to the existing error-free and uninterrupted guarantees.
Enterprise customers: LangChain will not use your data to train or improve large language models, making the restriction more explicit and comprehensive.
The updated terms introduce a new deployment model (BYOC) that allows customers to maintain their infrastructure while LangChain manages the platform through cloud APIs without direct access to customer data, providing an alternative to cloud-hosted and hybrid options. The terms now explicitly state that LangChain will not use customer data to train large language models, expanding the prior commitment beyond general product improvement. The warranty disclaimer has been revised to state that LangChain does not warrant the platform will be accurate or complete, in addition to existing disclaimers about error-free operation and uninterrupted service.
→ Evaluate whether the new BYOC deployment option better meets your organization's data residency or infrastructure control requirements compared to Cloud or Hybrid Deployments.
→ Review your data processing agreement with LangChain to confirm it reflects the expanded LLM training restriction and your chosen deployment model.
New deployment option allows customers to run LangSmith in their own infrastructure with LangChain managing it through cloud APIs without direct access to customer data.
Expanded commitment now explicitly states LangChain will not use customer data to train, develop, or improve large language models.
Revised to exclude warranties that the platform will be accurate or complete, in addition to existing error-free and uninterrupted disclaimers.
This change record describes what was added, removed, or modified in the document. Analysis reflects what the updated agreement states or permits. It does not constitute a legal determination about enforceability. Applicability may vary by jurisdiction. Methodology
Wording edits to this record that do not change what it claims. Changes to a claim are published as corrections.
LangChain has introduced a new deployment architecture (BYOC) and clarified data handling in Hybrid Deployments, which may affect how organizations evaluate data residency, compliance, and vendor risk. The explicit exclusion of large language model training …
Regulatory exposure, obligation change, escalation trigger, board-ready language, and recommended action for legal and compliance teams.
Unlock the full institutional analysis — InsightConductAtlas provides verified policy intelligence sourced directly from platform documents. All analysis is intended to support, not replace, legal and compliance review. Record CA-X-002279.
We read the privacy policies and terms of service of 38 AI platforms. Here is what they say about training, retention, arbitration, and lia…
Get the weekly coverage letter
Companies change their terms without announcing it. We capture governance documents, compare each new version with the last, and publish what changed, with the source text behind it. One email a week on what we found and what we could not reach.
Get alerted when this policy changes again, including what changed and why it matters.