AI Use Policy (Client and Public)

Last updated: 17 February 2026

We use human expertise first. Where appropriate, we may use carefully governed AI tools to support efficiency, consistency, and quality, always using safeguards designed to protect client confidentiality, privacy, and project requirements.

Our commitments to clients

• Choice and transparency: you can opt in or opt out of AI-assisted workflows, and you can set limits at any time.
• Your data stays yours: we do not use your content to train public AI models, or third-party AI models, and we do not permit third parties to do so without your explicit written consent.
• Security by design: we use access controls, encryption, vetted vendors, and retention practices proportionate to the sensitivity of your content.
• Human oversight: qualified professionals review and remain accountable for the final deliverables, particularly for specialist or high-stakes content.
• Ethical use: we aim to minimise bias and avoid misleading output, and we comply with applicable legal and contractual requirements.

How we may use AI

Depending on the service and your instructions, AI may be used to assist with specific tasks such as:

• Translation support (e.g., pre-translation or drafting) under controlled conditions and with appropriate human review.
• Terminology and glossary support (e.g., identifying candidate terms from non-sensitive sources).
• Quality checks (e.g., consistency, punctuation, formatting, and style conformance prompts).
• Formatting and layout assistance (e.g., producing consistent structure and templates).
• Internal training and process support via a ring-fenced internal AI assistant that uses approved internal documentation and role-based access controls, and not client confidential data unless expressly authorised.
• Traditional language technologies
  Traditional language technologies such as translation memories, terminology databases, and rule-based QA tools are not generative AI systems and may be used as part of standard translation workflows.

What we will not do

• We will not send confidential client content to public chatbots or consumer AI platforms.
• We will not use your content to train third-party AI models without your explicit written consent.
• We will not use AI in a way that breaches privacy, confidentiality, intellectual property rights, or applicable professional standards.
• We will not provide high-stakes deliverables solely generated by AI without appropriate human review and sign-off.
• We will not replace agreed human review or professional judgement with automated systems.

Your choices and controls

You can instruct us to:

• Opt out of AI-assisted workflows entirely for your projects.
• Approve limited, scoped AI uses only (e.g., allow formatting assistance but exclude translation drafting).
• Require redaction or de-identification measures for sensitive projects where technically and contractually feasible.
• Set specific retention and deletion preferences for project materials, where contractually and technically feasible.

Data protection and security

We use vetted, enterprise-grade tools and controls appropriate to the sensitivity of your content. These may include:

• Role-based access controls and least-privilege permissions.
• Supplier controls: where we use subcontractors, we require them to follow equivalent controls, and we prohibit the use of public chatbots on your content; compliance with these requirements forms part of our contractual terms.
• Encryption in transit and at rest, where supported by the platform.
• Approved tool lists, staff training, and blocking/monitoring of unapproved AI tools on corporate systems.
• Logging and auditability of AI-assisted workflows, where appropriate.
• Incident response procedures, including client notification, where required.

Questions

If you have questions about AI use on your projects or want an AI-free workflow, please contact your account/project manager. We will confirm your preferences in writing and configure the project accordingly.