Artificial Intelligence Systems

1. Definitions:
1.1 The following terms have the meanings set out below. Capitalised terms used but not defined in this clause have the meanings given to them in the Agreement.
"AI Laws" means all Laws which apply to the provision, marketing, deployment or use of AI Systems, including the EU AI Act (or any similar or equivalent laws in any other relevant jurisdiction) and Laws relating to personal data.
"AI Responsible Use Policy" means the policy governing use of AI systems by Customer’s suppliers, as updated or amended from time to time. 
"AI System" means a machine-based system that is designed to operate with varying levels of autonomy and that may exhibit adaptiveness after deployment, and that, for explicit or implicit objectives, infers from the input it receives, how to generate outputs such as content, predictions, recommendations or decisions that can influence physical or virtual environments.
“Customer Information" means any Customer Materials, Deliverables, Confidential Information, Output or other Customer data, information or materials;
"EU AI Act" means the Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024.
"Output" means any content, data, information, or other materials that are provided by the Supplier to Customer and created, generated or delivered through the use of an AI System.
2 The Supplier shall:
2.1 comply with all AI Laws  and must not do anything that would be a breach of or cause Customer to breach the AI  Laws; and
2.2 comply with Best Industry Practice in the implementation and operation of any AI System.
3 Where the Supplier intends to use an AI System for the provision of all or some of the Services and/or to process Customer Information, the Supplier must:
3.1 notify Customer prior to such use with details of the proposed AI System technology and functionality relevant to the provision of the Services including whether the AI System may use Customer Information as part of a prompt or instruction or in the course of training or refinement of such AI System or its underlying infrastructure; and
3.2 obtain Customer's prior written approval for such use.  
4 If Customer has approved the Supplier’s use of an AI System under clause 3, the Supplier must:
4.1 only use an AI System in accordance with any conditions for the approval given by Customer and the AI Responsible Use Policy.
4.2 conduct regular quality assurance checks in respect of the operation of the AI System and any Outputs used by the Supplier or provided to Customer in connection with the Services, including to determine the accuracy, reliability and suitability of such Outputs for the purpose for which they have been generated; 
4.3 implement all necessary measures to mitigate the likelihood of and remedy potential hallucinations, confidentiality and data privacy risks, risks of bias, discrimination and other applicable risks in Outputs in accordance with Best Industry Practice; and provide to Customer  promptly on written request details of the quality controls it employs; 
4.4 be able to suspend use of the AI Systems immediately on Customer’s instructions.  
4.5 not use a prohibited AI system or a high-risk AI system, as classified by the EU AI Act (or any AI Laws); and
4.6 not comingle, fuse or otherwise merge Customer Information with third-party data (or enable or permit a third-party to do such actions) including data owned by or in the possession of the Supplier (or its third-party providers) without Customer's express prior written consent.
5 Except as  expressly agreed by Customer in writing, the Supplier shall not (and shall not allow or permit any other party to) use, copy or otherwise utilise any Customer Information or Outputs for the purpose of training, developing or improving: any Supplier or third-party AI System; or its or any third party’s software and/or services more generally.
6 The Supplier shall provide information to Customer as may reasonably be required by Customer, relating to the use of AI Systems and/or any Outputs in the provision of the Services, in particular information on how the AI Systems are designed, developed and tested in such a way:
6.1 to ensure that its operation is sufficiently transparent to enable Customer to understand such AI Systems and associated risks;
6.2 that aligns with information sharing obligations imposed on the Supplier (or, where the Supplier is a downstream provider in respect of the AI System, those obligations imposed on the relevant upstream provider).
7 For the avoidance of doubt, the Outputs shall be included in the definition of Deliverables and Customer Confidential Information for all purposes under this Agreement.  In addition, the Supplier warrants to Customer that:
7.1 it has, or has procured that any relevant third-party provider has, incorporated all necessary transparency disclosures, labels, machine-readable detectable markings, and disclosures in the Outputs required to comply with the AI Laws and Best Industry Practice; and
7.2 no restrictions, conditions or limitations (whether contractual, legal, technical, or otherwise) arise from the use of such Outputs that would affect Customer's ability to use or modify the Deliverables or Services; and
7.3 it has all rights, permissions and consents to use the AI System (including by inputting data) and any Outputs in the course of delivering the Services and to enable Customer (and its customers and suppliers, where applicable) to receive the Services;
7.4 the use by Supplier of any input data, information or materials used in connection with an AI System in the course of delivering the Services, shall not breach the AI Laws, and/or infringe the rights of any third party.