Osprey Approach Artificial Intelligence Use Terms
Effective from 27 September 2026. Published 26 August 2026.
This policy forms part of the Osprey Approach Software Licensing & Services Agreement between Pracctice Limited and the Customer (Schedule H). Capitalised terms have the meanings given in that Agreement. In the event of any conflict, the Agreement prevails.
1. Application
1.1 This Schedule applies to any artificial intelligence, machine learning, generative artificial intelligence or similar functionality made available through or in connection with the Software (AI Functionality).
1.2 AI Output means any text, summary, analysis, recommendation, classification, draft, extraction, calculation or other output produced wholly or partly using AI Functionality.
1.3 AI Functionality forms part of the Software only to the extent that it is expressly made available by Pracctice. Nothing in this Agreement requires Pracctice to provide any particular AI Functionality unless expressly stated in the Osprey Approach System Features & Functionality Specification or the Commercial Schedule.
2. Customer Responsibility
2.1 The Customer is responsible for deciding:
whether AI Functionality is appropriate for its business and professional purposes;
which Users may use AI Functionality;
what information may properly be submitted to AI Functionality;
how AI Output is reviewed, verified and used; and
whether use of AI Functionality complies with the Customer’s legal, regulatory, professional and other obligations.
2.2 Use of AI Functionality does not transfer to Pracctice any responsibility owed by the Customer or its Users to clients, courts, regulators, counterparties or any other person.
2.3 The Customer remains responsible for all professional work, advice, communications, submissions, decisions and other actions undertaken using AI Functionality or AI Output.
3. Mandatory Human Review
3.1 AI Output must not be treated as authoritative, complete or independently verified.
3.2 The Customer must ensure that an appropriately competent person reviews and verifies AI Output before it is relied upon, communicated externally or used to make or support a material decision.
3.3 In particular, the Customer must independently verify:
legal propositions and legal analysis;
citations, authorities, legislation and case references;
factual statements;
financial or accounting information;
calculations;
dates and deadlines;
client or matter information;
documents intended for filing with a court, tribunal, regulator or public authority; and
any other information capable of materially affecting a client, matter or third party.
3.4 AI Output must not be used as a substitute for professional judgement, supervision or review required by law, professional regulation or the Customer’s own obligations.
4. Nature and Limitations of AI Output
4.1 The Customer acknowledges that AI Functionality may operate using probabilistic or statistical methods and may produce output which is inaccurate, incomplete, misleading, inconsistent, outdated, biased, unsupported by source material, based on incorrect assumptions or apparently authoritative despite being incorrect.
4.2 AI Functionality may generate references, citations, authorities, quotations, factual assertions or other information which does not exist or does not support the proposition for which it is produced.
4.3 Pracctice does not warrant that AI Output is accurate, complete, unique, current, legally correct, professionally appropriate or suitable for reliance without independent review.
4.4 Pracctice does not provide legal, accounting, regulatory or other professional advice through AI Functionality.
5. Customer Data, Confidentiality and Privilege
5.1 The Customer must not submit information to AI Functionality unless it is lawfully entitled to process that information for the intended purpose.
5.2 The Customer is responsible for determining whether Customer Data, confidential information, legally privileged material, Special Category Data, criminal-offence data or other sensitive information may lawfully and appropriately be submitted to AI Functionality.
5.3 Processing of Customer Personal Data through AI Functionality is governed by Schedule G.
5.4 The Customer must not submit Customer Data or other information obtained through the Software to an external or unsupported AI service by automated extraction, scraping, integration or other technical process unless authorised by Pracctice.
5.5 Pracctice will not use Customer Personal Data to train a general-purpose AI model except where the Customer has expressly instructed or agreed to such Processing, or the information has first been anonymised so that it is no longer Personal Data and cannot reasonably be associated with the Customer, its clients or Data Subjects.
6. Customer Instructions and Prompts
6.1 A prompt, instruction, configuration or other input provided by a User through AI Functionality is treated as an instruction from the Customer.
6.2 The Customer is responsible for the content and legality of such instructions and for ensuring that Users do not instruct AI Functionality to undertake Processing which the Customer itself would not lawfully or professionally be permitted to undertake.
6.3 Pracctice is not responsible for AI Output being inappropriate or incorrect because of an incomplete, inaccurate, ambiguous or inappropriate prompt, instruction, source document or other input supplied by the Customer or a User.
7. Prohibited Uses
7.1 The Customer and its Users must not use AI Functionality:
for any unlawful purpose;
to circumvent professional, regulatory or legal obligations;
to represent unverified AI Output as independently verified fact;
to knowingly create or submit false or misleading material;
to make a final professional judgement or material decision requiring human professional oversight without that oversight;
to gain unauthorised access to information, systems or another customer’s data;
to generate malicious code or facilitate unlawful interference with systems;
to discover, extract or reproduce system prompts, security controls, model configuration or other protected elements of the AI Functionality;
to circumvent usage, safety, security or access controls;
to reverse engineer or attempt to derive the underlying model, methods or architecture; or
to use AI Functionality or its outputs systematically to train, reproduce or develop a competing software product or AI model.
8. Third-Party AI Providers
8.1 AI Functionality may depend upon models, infrastructure or services supplied by third parties.
8.2 Where a third party processes Customer Personal Data on behalf of Pracctice in providing AI Functionality, that Processing is governed by Schedule G and the Subprocessor provisions contained there.
8.3 Pracctice may select, replace or cease using a third-party AI provider where reasonably required for technical, security, legal, regulatory, commercial or operational reasons.
8.4 Pracctice does not warrant the continued availability of any particular third-party model, model version or provider.
8.5 Where AI Functionality depends on a third-party service, Pracctice is not responsible for changes made independently by that provider to its underlying model, service or technical capability, except to the extent that Pracctice is required by this Agreement or applicable law to address the resulting effect on the Software.
9. Changes to AI Functionality
9.1 Pracctice may modify AI Functionality, including its underlying model or provider, prompts and system instructions, operating parameters, safety controls, supported use cases, functionality, technical limits and usage limits.
9.2 Pracctice may suspend or withdraw AI Functionality where reasonably necessary for security, safety, legal, regulatory, technical, commercial or operational reasons.
9.3 Suspension or withdrawal of AI Functionality does not constitute suspension or withdrawal of the Software as a whole.
9.4 Unless expressly stated otherwise in the Commercial Schedule, the availability of individual AI Functionality is not included in the Platform Availability target in Schedule B.
10. Usage Limits and Charges
10.1 Access to AI Functionality does not create an entitlement to unlimited AI usage.
10.2 Pracctice may impose reasonable usage, rate, volume or processing limits where necessary to manage service capacity, third-party costs, security, abuse or operational performance.
10.3 Where AI usage is subject to separate charges or included usage allowances, those commercial particulars will be stated in the Commercial Schedule or otherwise agreed with the Customer.
10.4 Changes in third-party AI provider charges may be passed through in accordance with the third-party pricing provisions of Clause 5.
11. Intellectual Property
11.1 Customer Data submitted to AI Functionality remains Customer Data and is governed by the provisions of this Agreement relating to Customer Data.
11.2 Pracctice retains all rights in the Software, AI Functionality, system prompts, configurations, methods, workflows, templates and other proprietary elements provided by Pracctice.
11.3 To the extent that Pracctice has the right to do so, the Customer may use AI Output produced through its authorised use of the Software for its own business purposes.
11.4 Because of the nature of AI systems, Pracctice does not warrant that AI Output is unique, capable of intellectual property protection, dissimilar from output provided to another person or free from the need for further review or modification before use.
11.5 The Customer is responsible for determining whether an AI Output may lawfully be used, reproduced, published or supplied to another person.
12. Records and Regulatory Compliance
12.1 The Customer is responsible for maintaining any records of AI use, prompts, review or approval required by its professional regulator, applicable law or its own governance requirements.
12.2 Pracctice may retain technical and audit information concerning use of AI Functionality where reasonably required for security, support, service operation, investigation of misuse, regulatory compliance and improvement of the Software, subject to Schedule G where that information contains Personal Data.
12.3 The Customer will cooperate with reasonable requests by Pracctice concerning investigation of suspected misuse of AI Functionality.
13. Suspension
13.1 Pracctice may suspend a User’s access to AI Functionality where it reasonably believes that:
this Schedule has been materially breached;
use presents a security or operational risk;
continued use may be unlawful;
continued use creates a material risk to Pracctice, another customer or a third party; or
suspension is reasonably required by a third-party provider or competent authority.
13.2 Where reasonably practicable, suspension under this Section will be limited to the affected AI Functionality or User.
13.3 Suspension of AI Functionality under this Section does not relieve the Customer of its payment obligations under the Agreement.
14. Responsibility for Use and Claims
14.1 Subject to Pracctice’s obligations under applicable law and this Agreement, the Customer is responsible for losses, claims and consequences arising from its decision to rely on AI Output without the review required by this Schedule, unlawful or unauthorised Customer inputs or instructions, use of AI Output contrary to law or professional regulation, failure to verify material AI Output, or modification or use of AI Output after it has been supplied by the Software.
14.2 The Customer will indemnify Pracctice against third-party claims, regulatory costs and reasonable professional costs arising from the matters in Section 14.1, except to the extent caused by Pracctice’s breach of this Agreement or applicable law.
14.3 The application of the exclusions, limitations and indemnities in this Schedule is subject to Clause 9.
15. Relationship with the Agreement and Acceptable Use Policy
15.1 This Schedule forms part of the Agreement and is incorporated into the Acceptable Use Policy by reference.
15.2 A breach of this Schedule is also a breach of the Acceptable Use Policy.
15.3 The Customer’s professional, regulatory and legal responsibilities are not reduced by the use of AI Functionality.
15.4 Nothing in this Schedule extends the functionality, availability, warranty or Service Level obligations otherwise undertaken by Pracctice under this Agreement.
15.5 In the event of inconsistency between this Schedule and another provision relating specifically to AI Functionality, this Schedule will apply, subject always to any provision of the Main Agreement which expressly states that it has precedence.
AMENDMENT TO THESE AI USE TERMS
Published 30 August 2026. Effective from 30 September 2026.
This amendment forms part of these AI Use Terms from its effective date and is published by website posting in accordance with the Main Agreement. Until the effective date, the existing terms apply unchanged.
1. AI systems and platform access channels
The use of customer API access and the Osprey Insights Database in connection with artificial intelligence systems is governed by the API Terms, including the amendment to the API Terms published 30 August 2026 and effective from 30 September 2026. In summary, and without limiting the API Terms: those access channels may not be used to train, fine-tune, or ground AI systems on the structure, schemas, or systematically extracted contents of the Software; to develop or operate functionality that replicates or substitutes for any part of the Software; or to operate an intermediary interface (including AI agents) through which users interact with Customer Data in place of the Software. Using AI tools to analyse reports, dashboards and outputs produced from the Osprey Insights Database is operational use and is permitted. If there is any conflict between this section and the API Terms, the API Terms prevail in respect of those access channels.
