<aside> 📄
Download this document as PDF
Conduct AI Artificial Intelligence Annex v1.1.pdf
</aside>
Version: 1.1 Effective Date: March 17, 2026
This Annex ("AI Annex") sets out Conduct AI's specific commitments regarding the use of artificial intelligence in delivering the Platform. It supplements the General Terms of Service ("ToS") and, in the event of any conflict with the ToS on AI-specific matters, this Annex shall prevail.
In this Annex, the following terms have the meanings set out below. Capitalised terms not defined here have the meanings given in the ToS.
1.1 "Provider" means a third-party company whose large language model or generative AI services are integrated into the Platform for the purpose of analysing Customer Data and generating Output. An AI Provider is identified at the company level; individual models, model versions, and releases within an AI Provider's product family do not constitute separate AI Providers.
1.2 "Approved Provider List" means the list of AI Providers currently approved by Conduct AI for use in the Platform, as set out in Appendix 1 and updated from time to time in accordance with Clause 4.1.
1.3 "Model-Level Update" means the release, substitution, or upgrade of any model version, variant, or iteration within an existing AI Provider's product family. A Model-Level Update does not constitute the introduction of a new AI Provider and does not trigger the notification requirement under Clause 4.1.
2.1 Conduct AI's commitment. Conduct AI will not use Customer Data to train, fine-tune, retrain, or otherwise adapt any machine learning or AI model, whether operated by Conduct AI or any third party.
2.2 AI Provider commitment. Each AI Provider on the Approved Provider List is subject to a contractual commitment prohibiting it from using Customer Data to train or improve its models. Conduct AI will not add a new AI Provider to the Approved Provider List unless an equivalent commitment is in place. Evidence of such commitments is available on written request.
3.1 Provider-level framework. Conduct AI manages its AI supply chain at the AI Provider level. Accordingly:
(1) Conduct AI will notify the Customer in writing at least 15 days before integrating an AI Provider not currently on the Approved Provider List. The notification will include the provider's name, the nature of services, and the applicable no-training commitment. The Customer may raise reasonable objections in writing. If the objection cannot be resolved and Conduct may not be able to provide the Service without the new AI Provider, the Customer may terminate the affected Order Form on 30 days' written notice and receive a pro-rata refund of prepaid Fees for the unused term.
(2) Conduct AI may integrate or upgrade to any Model-Level Update of an AI Provider being on the Approved Provider List at any time and without prior notice to the Customer (e.g. GPT 3.5 to GPT 4o).
3.2 Approved Provider List. The current Approved Provider List is set out in Appendix 1.
4.1 Output accuracy. The Platform uses probabilistic AI. Generated Output may contain inaccuracies or errors ("Hallucinations") and constitutes a recommendation only. The Customer is responsible for reviewing and validating all Generated Output before reliance or use in any business process. Conduct AI does not warrant accuracy, completeness, or fitness for purpose of any Generated Output.
4.2 Provider carve-out. Conduct AI will not be liable to the Customer for any loss, damage, service interruption, or inaccuracy in Generated Output to the extent directly caused by the acts, omissions, or technical failures of an AI Provider including model errors, hallucinations originating from the underlying model, service outages, or unilateral changes to an AI Provider’s APIs or policies. For the avoidance of doubt, Conduct AI remains responsible for its own acts and omissions in the configuration and orchestration of AI Provider services.