The following Service Specific Terms apply to the specific Databricks Services indicated below. In the event of a conflict or inconsistency between the terms of these Service Specific Terms and the terms of the Databricks Master Cloud Services Agreement or other agreement with us governing your use of the Databricks Services (the “Agreement”) the terms and conditions of these Service Specific Terms apply except as noted in Section 1.4 below. Capitalized terms used herein but not defined herein shall have the meanings set forth in the Agreement (or if not therein defined, shall refer to the term in such Agreement that is closest in meaning to such term).
You may sign up to receive updates of these Service Specific Terms here.
- Universal Terms (applicable to all Databricks Services).
- Your use of the Platform Services must at all times comply with the Acceptable Use Policy.
- You must comply with the current Documentation as posted by us and updated by us from time to time on https://www.databricks.com/documentation (or such other website as we provide).
- If you process the personal data of Authorized End Users or other identifiable individuals in your use of a Databricks Service, you are responsible for providing legally adequate privacy notices and obtaining necessary consents for the processing of such data. You represent to us that you have provided all necessary privacy notices and obtained all necessary consents. You are responsible for processing such data in accordance with applicable law.
- Your use of any Databricks Service described below is subject to these Service Specific Terms notwithstanding anything to the contrary in an Agreement, provided that if your Agreement already contains terms that specifically cover the indicated features below (such terms, “Overriding Terms”) (e.g., you entered into the Databricks Master Cloud Services Agreement at https://www.databricks.com/legal/mcsa from 28 March 2022 onwards), the terms in your Agreement will apply and the terms for that specific feature will be considered superseded by your Agreement; if your Agreement limits the applicability of these Service Specific Terms, you are not authorized to enable or access a Databricks Service covered by these Service Specific Terms unless such feature is covered by Overriding Terms.
- You may perform benchmarks or comparative tests or evaluations (each, a “Benchmark”) of the Platform Services and may disclose the results of the Benchmark other than for Beta Services. If you perform or disclose, or direct or permit any third party to perform or disclose, any Benchmark of any of the Platform Services, you (i) will include in any disclosure, and will disclose to us, all information necessary to replicate such Benchmark, and (ii) agree that we may perform and disclose the results of Benchmarks of your products or services, irrespective of any restrictions on Benchmarks in the terms governing your products or services.
- Databricks may update these terms from time to time. Updates of these terms shall be deemed effective: (a) for new products and services, upon posting of the updated terms here; and (b) for updates to these terms that were previously posted (i) for Monthly PAYG Service customers (i.e., customers who are not on committed usage contracts), 30 days after such update is made to these Service Specific Terms and notified as described herein or (ii) for all other customers, upon renewal of the applicable Order Form. Databricks will provide notice to anyone who has subscribed to receive an update to these terms as indicated in the introduction above and may additionally provide notice to administrative users of the Platform Services.
- Definitions (applicable to all Platform Services).
- “Cloud Environment” of a party means the cloud or other compute or storage infrastructure controlled by such party and utilized under the Agreement.
- “Compute Plane” means the portion of the applicable Cloud Environment where compute resources of the Platform Services are deployed during use of the Platform Services for the primary processing of Customer Data. In the case of Serverless Compute, the Compute Plane is within the Databricks Cloud Environment (the “Databricks Compute Plane”). In all other cases, the Compute Plane is within the Customer Cloud Environment (the “Customer Compute Plane”), and in such cases the processing activity results in fees being charged directly to Customer by the Cloud Service Provider. For the avoidance of doubt, none of the terms “Compute Plane”, “Customer Compute Plane” or “Databricks Compute Plane” include Customer’s cloud storage. In certain Databricks agreements and Documentation, the Compute Plane may be interchangeably referred to as the “Data Plane”.
- “Databricks Control Plane” means the elements of the Platform Services residing within the Databricks Cloud Environment, other than the Databricks Compute Plane, including without limitation the user interface of the Platform Services.
- "Serverless Compute” means a Platform Service where the Compute Plane is located in Databricks’ Cloud Environment rather than in Customer’s Cloud Environment.
- Platform Services - General.
- Deployment and Responsibilities.
- Generally. Customer acknowledges that the Platform Services are implemented in a manner that divides the Platform Services between the Customer Cloud Environment and the Databricks Cloud Environment, and that accordingly each party must undertake certain technical and organizational measures in order to protect the Platform Services and the Customer Content. Without limiting the foregoing, Customer acknowledges and agrees that (1) in order to utilize the current Platform Services other than Serverless Compute, Customer must have an account with the Cloud Service Provider; (2) Databricks does not host the Customer Cloud Environment into which certain parts of the Platform Services may be deployed or the Systems in which your Customer Data may be stored (e.g., an AWS S3 bucket); (3) while certain Customer Data may in some cases be present within the Databricks Cloud Environment of the Platform Services (e.g., within the Customer Results), Databricks’ current Platform Services are not designed to archive or permanently retain Customer Data (e.g., they are intended to provide an environment to facilitate Customer’s processing of Customer Data by permitting Customer to execute Customer Instructional Input and view Customer Results); and (4) Databricks’ current Platform Services do not provide backup services or disaster recovery to enable recovery of Customer Data. Accordingly, and without limiting the foregoing, Databricks is not responsible for any loss, destruction, alteration, or corruption of Customer Content, except to the extent caused by the gross negligence or willful misconduct of Databricks.
- Your Responsibilities. You acknowledge and agree that you are responsible for (1) protecting the security of all your credentials used to access the Platform Services (with Databricks also responsible for taking adequate steps to protect Customer credentials to the extent such credentials are within the control of Databricks); (2) securing the Customer Cloud Environment and any Customer System (with such steps to include without limitation the regular rotation of access keys and other industry standard steps to preclude unauthorized access); (3) backing up Customer Instructional Input (e.g., via Github or other third party System); (4) backing up and securing Customer Data under Customer’s control within the Customer Cloud Environment or other Customer controlled System (e.g., by turning on versioning and encryption within AWS S3); (5) any security or other issues resulting from any Customer Instructional Input; and (6) managing and paying the charges associated with your usage of the Customer Cloud Environment (e.g., compute and storage fees); and Customer expressly assumes the risks associated with the foregoing responsibilities set forth in this Section.
- Databricks Responsibilities. Databricks acknowledges and agrees that, as between the parties and except to the extent caused by the action or intentional or negligent inaction of you or your Authorized Users, including without limitation any customizations or configurations of the Platform Services by you or anything specified to be your responsibility above, Databricks is primarily responsible for: (1) the operation (excluding to the extent such operation is directed by the Customer Instructional Input) of the Databricks Cloud Environment (including the user interface of the Platform Services, the Databricks Compute Plane with respect to Serverless Compute, and the portion of the Platform Services within the Databricks Control Plane in which the Customer Instructional Input and Customer Results are held until deleted by you) and, with respect to Platform Services other than Serverless Compute, the Databricks software that operates the computing resources in the Customer Compute Plane; and (2) implementing reasonable technical and organizational measures designed in accordance with ISO/IEC 27001:2013 or equivalent/greater standard to protect the security of the foregoing. Additionally, while it is your responsibility to back up Customer Instructional Input, Databricks will, at your reasonable request, provide commercially reasonable assistance with recovery efforts where reasonably possible.
- Platform Services other than Serverless Compute. With respect to your use and Databricks’ provisioning of Platform Services other than Serverless Compute, including without limitation All Purpose Compute, Jobs Compute (including Jobs Light Compute) and SQL Compute using Classic SQL Endpoints, the Compute Plane is deployed within the Customer Cloud Environment. Notwithstanding the foregoing, Databricks may in the future add additional Platform Services that may be subject instead to other Service Specific Terms by indicating in the Databricks release notes at the time of release that such Platform Services are subject to additional Service Specific Terms.
Serverless Compute (part of the Platform Services; in Public Preview on Databricks on AWS as of the date of these Service Specific Terms). Databricks Serverless Compute, including without limitation Serverless SQL Endpoints, operates by Databricks deploying the Compute Plane into the Databricks Cloud Environment instead of the Customer Cloud Environment. Accordingly, your use of and/or our provision of Databricks Serverless Compute may conflict with or be inconsistent with certain terms you have previously agreed to with Databricks in an Agreement (including without limitation, the terms relating to your deployment or the respective responsibilities of the parties in the terms of service and/or data processing addendum governing the Platform Services, as well as any security schedule or other terms), particularly as such terms may relate to our security obligations or our rights and responsibilities regarding data processing.
- Unless expressly permitted in an Order Form, Customer may not use Databricks Serverless Compute to process Cardholder Data or PHI.
- Downloadable Services.
- Databricks may make available to you certain Databricks Services as software from time to time in a downloadable manner (“Downloadable Services“). Unless expressly stated otherwise at the time of download or as otherwise agreed by Databricks, Downloadable Services may only be used as a part of the Databricks Services. You are granted a non-exclusive, royalty-free right and license to use and copy during the term of the Agreement the Downloadable Services solely as necessary to enable your use of the Databricks Services.
- Beta Services.
- “Beta Service” means any Databricks Service (or feature of a Databricks Service) that is clearly designated as “beta”, “experimental”, “preview” or similar, that is provided prior to general commercial release, and that Databricks at its sole discretion offers to Customer, and Customer at its sole discretion elects to use.
- If you elect to receive any Beta Services offered by Databricks, you agree that, in addition to adhering to all other restrictions generally applicable to your use of the Platform Services under the Agreement and any requirements set forth by Databricks in writing regarding the particular Beta Services, you will not use such Beta Services for production workloads or for any mission critical work, and that you will not use sensitive data (e.g., PHI or Cardholder Data) in conjunction with such Beta Services unless explicitly permitted in an Order Form. For the avoidance of doubt, information pertaining to the Beta Services constitutes Databricks Confidential Information.
- BETA SERVICES ARE PROVIDED “AS-IS” AND WITHOUT WARRANTY OF ANY KIND. CUSTOMER BEARS THE RISK OF USING THE BETA SERVICES. DATABRICKS AND ITS SUPPLIERS GIVE NO EXPRESS OR IMPLIED WARRANTIES, GUARANTEES OR CONDITIONS RELATED TO THE BETA SERVICES. TO THE EXTENT PERMITTED UNDER YOUR LOCAL LAWS, DATABRICKS AND ITS SUPPLIERS EXCLUDE THE IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT.
- Databricks Container Services (part of the Platform Services).
- As part of the Databricks Container Services, Databricks may provide a sample stub container file (a “Sample Container”) that you may (but are not required to) use to create a custom container file to use with Databricks Container Services (a “Modified Sample Container”). Databricks grants you a limited, non-exclusive right and license to use and modify the Sample Container by inserting your own code into the Sample Container to create a Modified Sample Container. The Sample Container may contain libraries that are subject to open source licenses. It is your obligation to review and comply with any such licenses prior to the creation of the Modified Sample Container.
- You may not:
- upload to the Platform Services or otherwise make available for use with Databricks Services a container file (including a Modified Sample Container) (together, a “Custom Container”) or other code that contains any viruses, worms, time bombs, Trojan horses and other harmful or malicious code, files, scripts, agents or programs designed to impede the use of the Databricks Services or any other systems, whether or not connected to the Databricks Services;
- include in a Custom Container any code for which you do not have the necessary right or license or that contains any code that could subject Databricks Services (including the Databricks Container Services) to any term that seeks to require Databricks to make any Databricks Services code available in source code form or which may impose any other obligation or restriction with respect to Databricks’ intellectual property rights; or
- attempt to disable or interfere with any technical limitations contained within Databricks Container Services.
- You grant Databricks a worldwide, non-exclusive royalty free right and license to use, reproduce and make derivative works of the Custom Container solely as necessary to provide Databricks Container Services.
- Delta Sharing (part of the Platform Services; in Preview on Databricks on AWS as of the date of these Service Specific Terms).
- Delta Sharing allows a Databricks Authorized User to share data outside of the Authorized User’s Customer organization. Accordingly, Customer’s use of and/or Databricks’ provision of Delta Sharing may conflict with or be inconsistent with certain terms you have previously agreed to with Databricks in an Agreement (including, without limitation, terms relating to your deployment or the respective responsibilities of the parties in the terms of service and/or data processing addendum governing the Platform Services, as well as any security schedule or other terms), particularly as such terms may relate to our security obligations or our rights and responsibilities regarding data processing.
- Additional Terms Relating to your Deployment.
- Additional Definitions. The following additional terms apply to any use of Delta Sharing (as defined below).
- "Data Recipient" means a person who receives Customer Data from a Data Provider through Delta Sharing. A Data Recipient may be another Databricks customer or may be a third party entity that is not a Databricks customer (“Non-Customer Recipient”).
- "Data Share" means an individual set of Shared Data.
- "Data Provider" means a person who initiates the share of the Customer Data from the Platform Services.
- "Delta Sharing" means the portion of the Databricks Platform Services that enable a Data Provider to share Customer Data with a Data Recipient, as further described in the Documentation.
- "Shared Data" means Customer Data (either Customer’s Customer Data or the Customer Data of another Databricks customer) that is shared via Delta Sharing.
- Generally. The Platform Services include functionality that lets Customer, at its option and in its sole discretion, provide Data Recipients outside of its Databricks workspace(s) with read-only access to Shared Data or receive Shared Data from other Data Providers.
- When Customer is the Data Provider. Customer may, at its option and in its sole discretion and subject to the rest of the Agreement, use Delta Sharing to grant one or more Data Recipients read-only access to some or all of its Customer Data via one or more Data Shares. Customer:
- is solely responsible for ensuring that it is entitled to share any Shared Data with any Recipient and that any such Data Share is in compliance with all applicable laws.
- acknowledges and agrees that (i) Data Recipients will have ongoing access to the Shared Data with the permissions set by Customer until the applicable Data Share is disabled; (ii) nothing in Delta Sharing prevents a Data Recipient from making a permanent copy of the Shared Data; (iii) Databricks has no control over, and will have no liability for, the acts or omissions of any Data Recipient with respect to the Shared Data; and (iv) Customer remains responsible for all Customer Data in accordance with the Agreement, irrespective of whether Customer or Data Recipient is in control of such data.
- as between Databricks and Customer, acknowledges and agrees that it is responsible for any fees (including fees charged by a Cloud Service Provider) associated with the access to and use of any Shared Data by a Data Recipient.
- represents that it has the right to share with Databricks any personal data associated with any Data Recipients that it shares with Databricks.
- acknowledges and agrees that it is fully responsible for any actions or inactions of a Data Recipient with respect to the Shared Data and will, subject to the conditions of indemnification set forth in the Agreement, but notwithstanding any liability cap or disclaimer or exclusion of damages therein (or otherwise agreed between the parties), defend all Databricks Indemnitees from and against any claims, demands, suits or proceedings made by a third party including without limitation a Data Recipient, and will indemnify each Databricks Indemnitee from and against any damages, attorneys’ fees and costs finally awarded against a Databricks Indemnitee, arising from or related to any Shared Data provided to any Data Recipient through the Delta Sharing. “Databricks Indemnitees”, if no equivalent term is defined in the Agreement, shall mean Databricks and its Affiliates and its and each of their officers, employees, directors, and agents.
- When Customer is the Data Recipient. By accessing a Data Provider’s Shared Data as a Data Recipient, Customer represents that it has been authorized to access the Data Share(s) provided to it by the Data Provider and acknowledges that (1) Databricks has no liability for such Shared Data or Customer’s use of such Shared Data, and (2) Databricks may collect information about Customer’s use of and access to the Shared Data (including identifying Customer and any individual who accesses the Shared Data using the credential file in connection with such information) and may share it with the applicable Data Provider.
- Restrictions on Shared Data; Suspension. Unless set forth in an Order Form, Customer may not include any PHI (as defined by the Health Insurance Portability and Accountability Act) or Cardholder Data (as defined by PCI-DSS) in any Shared Data. Customer acknowledges and agrees that Databricks may, but is not required to, disable a given Data Share or Customer’s use of the Delta Sharing or access to a Data Share if Databricks reasonably suspects or becomes aware of a claim that a Data Share may violate applicable law or may harm Databricks, Customer, Recipient or other third party.
Last Updated March 28, 2022. For earlier versions, please send a request to [email protected] (with “Service Specific Terms Request” in the subject).