AI Addendum

Bonterms Standard AI Addendum v2.0 (Attachment Version) — adopted by Continuum API

This AI Addendum governs Customer use of AI-powered features within the Continuum API platform and is based on the Bonterms Standard AI Addendum v2.0 (CC BY 4.0). It attaches to and is incorporated into the Cloud Terms. In case of conflict, this Addendum controls.

AI Features currently in the Continuum API platform include: AI-assisted sequence generation, AI-powered subject-line and copy suggestions, and automated campaign scheduling recommendations. Non-generative machine-learning components (e.g., deliverability scoring models) are not covered by this Addendum.

Model Provider: Anthropic (claude.ai). See Subprocessors for details.

1. Definitions

Terms defined in the Cloud Terms control where equivalent language exists. Key definitions:

• AI Features: All generative artificial intelligence, large language model (LLM) or similar features that generate content in response to Input, excluding non-generative machine-learning features.
• Input: Data or materials Customer submits to AI Features.
• Output: Content generated by AI Features in response to Input.
• Model Provider: Third-party AI model or inference service providers (currently Anthropic).
• Retention Period: 30 days.

2. AI Features

2.1 Provider identifies AI Features in Documentation or upon written request.

2.2 New AI Features. When Provider adds New AI Features: (a) Provider will give reasonable prior written notice before enabling them by default; (b) Customer may decline or disable them through administrative controls; and (c) disabling New AI Features will not materially reduce core non-AI functionality that existed before their introduction.

3. Data Handling

3.1 No Training. Provider will not access, use, or retain Input or Output to train, retrain, or fine-tune AI Features, any Model Provider models, or other AI or machine-learning models.

3.2 Operational Monitoring. Provider may retain Input and Output for no longer than the Retention Period solely to detect, prevent, and remediate abuse, misuse, security incidents, malfunctions, or abnormal performance. Such data must be (a) logically separated from production data, (b) accessible only to authorized personnel on a need-to-know basis, and (c) automatically deleted upon Retention Period expiration unless required by law or needed for ongoing investigations.

3.3 Inference and Processing. Provider represents and warrants: (a) Tenant Isolation — caching, indexing, or retrieval systems are logically separated on a per-customer basis; Customer's Input/Output will not populate shared context windows accessible to other customers; and (b) Ephemeral Processing — except for Operational Monitoring or Customer election, processing is transient and Input/Output is not retained beyond the time necessary for session generation.

4. Intellectual Property

4.1 Input. Input is deemed Customer Data for all Cloud Terms purposes, including intellectual property ownership.

4.2 Output. As between the parties, Customer owns the Output. Provider assigns to Customer all of Provider's right, title, and interest (if any) in and to the Output. Provider retains its intellectual property rights in the Cloud Service. Customer acknowledges that Output may not be unique; different customers may receive similar or identical Output.

5. Output Indemnification

5.1 Copyright Indemnity. Provider-Covered Claims include third-party claims alleging that Customer's use of Output infringes a third-party copyright.

5.2 Exclusions. Provider's indemnification does not apply to Output Claims resulting from Customer (a) failing to use AI Features per Documentation or safety filters, (b) providing infringing Input or Input designed or likely to generate infringing Output, or (c) modifying Output or combining it with other materials where infringement results from the modification or combination.

5.3 Liability Cap. Output Claims are subject to the Cloud Terms' general limitation of liability but not the consequential damages waiver.

6. AI Warranties and Disclaimers

6.1 Safety Measures. Provider will implement commercially reasonable technical and organizational measures to mitigate harmful or illegal Output generation.

6.2 Output Disclaimer. Customer acknowledges that due to the probabilistic nature of generative AI, Output may contain errors, inaccuracies, or biases. Provider makes no accuracy or suitability warranties. Customer is solely responsible for independently reviewing and verifying Output before relying on, publishing, or otherwise using such Output.

6.3 No High-Risk AI. Provider warrants AI Features do not constitute a "High-Risk AI System" as defined under the EU AI Act, unless explicitly disclosed in Documentation.

7. AI-Specific Use Restrictions

In addition to the Cloud Terms' usage rules, Customer will not use AI Features or Output to:

• Train third-party AI models that compete with Provider or Model Providers;
• Represent Output as Provider-approved;
• Represent Output as wholly human-generated;
• Make automated decisions with legal or significant effects on individuals without appropriate human review and legal compliance;
• Violate Model Provider acceptable use policies; or
• Engage in practices prohibited under the EU AI Act or equivalent AI law.

8. Model Providers

8.1 Customer acknowledges AI Features may utilize Model Providers. Provider maintains a current Model Providers list at /legal/subprocessors and notifies Customer of changes.

8.2 Customer authorizes Provider to use Model Providers to process Input and Output.

8.3 Model Provider use does not change Provider's responsibility for the Cloud Service and AI Features.

8.4 Unless expressly agreed, Customer is not required to enter separate Model Provider agreements.

8.5 Provider ensures Model Providers are subject to contractual obligations regarding Input and Output that are materially consistent with Section 3 (Data Handling).

8.6 Temporary Suspension. Provider may suspend or limit AI Features due to Model Provider unavailability, applicable law, or material security/safety risk. Provider will notify Customer and use commercially reasonable efforts to restore services.

9. General Terms

9.2 Relationship to Cloud Terms. This AI Addendum is incorporated into the Cloud Terms. In conflicts between this Addendum and the Cloud Terms, this Addendum controls. Otherwise, Cloud Terms terms apply.