AI Terms

Last Revised: September 29, 2026

1. Scope

These AI Terms govern Customer’s use of the AI Features. They form part of the Currents Terms of Service at https://currents.dev/tos (“TOS”). Capitalized terms not defined here have the meaning given in the TOS. If these AI Terms conflict with the TOS, these AI Terms control for the AI Features, except that an Order prevails over these AI Terms.

2. Definitions

“AI Features” means the features of the Services that use generative AI models to analyze Customer Content or to respond to prompts from Permitted Users.

“AI Provider” means a third party whose model or hosted inference service Currents uses to provide the AI Features, whether Currents accesses it directly or through a hosting, gateway or routing service (for example Amazon Bedrock, Microsoft Azure or Google Cloud Vertex AI).

“AI Input” means the Customer Content and prompts that Currents sends to an AI Provider to provide the AI Features.

“AI Output” means the content that the AI Features return to Customer.

3. Control of the AI Features

Administrators of Customer’s organization control whether the AI Features are enabled for the organization, and can enable or disable them at any time through the Services. While the AI Features are enabled, Customer instructs Currents to send AI Input to the AI Providers and confirms that Customer is permitted to share its Customer Content with the AI Providers. Disabling stops new AI Input from being sent; requests in progress may fail, and AI Input already sent remains subject to Section 5.

4. AI Input

AI Input consists of the Customer Content that Currents selects as relevant to the request, together with any text a Permitted User enters into an AI Feature. Customer controls the Customer Content it sends to the Services and is responsible for it, and shall disable the AI Features if its Customer Content contains data that Customer is not permitted to share with the AI Providers.

5. AI Providers

The current AI Providers are listed at https://currents.dev/sub-processors. Currents may select, add, remove or replace AI Providers, models, and the hosting, gateway or routing services through which it accesses them, subject to the notice procedure in the DPA where personal data is concerned.

Currents engages AI Providers under terms that (i) prohibit the AI Provider from using AI Input or AI Output to train or improve its models, and (ii) limit the AI Provider’s retention of AI Input and AI Output to the period needed to provide the service and to detect abuse, as described in the Currents documentation, unless otherwise agreed in an Order.

Once AI Input is sent, it is processed on the AI Provider’s systems under the AI Provider’s terms. Currents does not control those systems and, except for Currents’s obligations for personal data under the DPA and to the extent permitted by applicable law, is not responsible for the acts or omissions of the AI Providers. An outage or failure of an AI Provider is not a breach of the TOS by Currents.

6. Improving the Services

The use of AI Input and AI Output to develop and improve the Services, including the AI Features, is governed by the “Improving the Services” paragraph in Section 5 of the TOS.

7. AI Output

AI Output is Customer Content and, as between the parties, Customer owns it, subject to any rights of third parties in material it contains. Generative AI models produce output that may be inaccurate, incomplete or misleading, and that may be the same as or similar to output produced for other customers. Customer shall review AI Output before relying on it and is solely responsible for its use of AI Output. AI Output is not professional advice. Currents’s obligation under Section 21(ii) of the TOS does not apply to AI Output. Section 21(iv) of the TOS applies to AI Output only to the extent the claim arises from AI Input or from Customer’s use of the AI Output.

An AI Feature may take actions in Customer’s Account at the request of a Permitted User. It does so with that Permitted User’s permissions, and the action is treated as taken by that Permitted User. Customer is responsible for reviewing an action before confirming it where the AI Feature asks for confirmation.

8. Use of the AI Features

Customer shall not use the AI Features:

  • (i) to attempt to extract, reproduce or reverse engineer the models, system prompts or configuration used by the AI Features,
  • (ii) to develop or train a model that competes with the AI Features or with the AI Providers’ models,
  • (iii) in breach of the AI Providers’ usage policies that Currents refers to in its documentation,
  • (iv) as part of a system that makes decisions with legal or similarly significant effects on natural persons, or for any use that applicable law prohibits or classifies as high-risk for AI systems,
  • (v) to present AI Output as created by a person, or as reviewed or endorsed by Currents, or
  • (vi) in a manner that violates Section 7 or Section 14 of the TOS.

9. Warranties and Preview Features

The AI Features are excluded from the Services Warranty in Section 18 of the TOS and from any service level or uptime commitment.

Currents may change, suspend or discontinue any AI Feature, model or AI Provider at any time. AI Features that Currents labels as beta, preview or experimental are provided for evaluation and may be changed or withdrawn without notice. Currents’s aggregate liability arising from such features shall not exceed the lower of one thousand U.S. dollars (US$1,000) and the fees actually received by Currents from Customer in the twelve (12) months immediately preceding the first event giving rise to a claim.

10. Retention

Currents retains AI Output and Permitted Users’ interactions with the AI Features for the periods set out in the data retention table at https://docs.currents.dev/resources/data-privacy/data-retention.

11. Changes

Currents may update these AI Terms in accordance with Section 24 of the TOS. During a paid Term, an update will not reduce Currents’s commitments in Sections 3 and 5 of these AI Terms without Customer’s consent.

If you have any questions (or comments) concerning these AI Terms, please send us an email to support@currents.dev.