Responsible AI Policy

Guiding North HR uses artificial intelligence carefully, in limited situations and with Derek Carter accountable for the outcome. AI may support the work, but it does not replace professional judgement, human conversation or responsibility.

Derek may use approved tools such as Microsoft Copilot, ChatGPT and Claude for:

• calculations and numerical checks;
• analysis of genuinely anonymised or aggregated data;
• identifying patterns or trends in non-identifiable information;
• summarising general, non-confidential research or business information; and
• supporting administrative efficiency where no confidential or identifiable client information is exposed.

Guiding North HR does not use AI:

• to make solely automated decisions about employees, workers, candidates or clients;
• to decide or recommend final outcomes in recruitment, performance, discipline, grievance, redundancy, dismissal, pay or other matters affecting a person's rights or livelihood;
• to produce client legal documents or as a substitute for advice from a qualified legal professional;
• to process identifiable or confidential client, employee or enquiry information in general-purpose AI tools;
• to create deceptive content, impersonate people or conceal a material AI-generated decision; or
• to rely on an AI output without appropriate human checking.

Information must be genuinely anonymised before an approved AI analysis. Names, email addresses, employee numbers, precise job details, rare circumstances and other identifying features must be removed or generalised so that Derek and the AI provider cannot reasonably identify the person from the information supplied.

Removing a name alone is not anonymisation if the remaining details can still be linked to a person. That is usually pseudonymisation, and the information remains personal data. If there is doubt, the information will not be entered into the AI tool.

Before using an AI service for business work, Derek will consider its purpose, account type, privacy and security settings, contractual terms, data retention, model-training controls, subprocessors and international transfers. Only tools and configurations suitable for the intended low-risk use will be approved.

A newly available AI feature is not automatically approved simply because it appears inside software Guiding North HR already uses.

Derek remains responsible for checking relevance, accuracy, calculations, assumptions, bias and context. AI outputs can be incomplete, outdated or confidently wrong. Material conclusions are tested against reliable sources and Derek's professional experience before use.

AI will not be used in a way that unfairly discriminates, profiles people or creates unjustified adverse effects. If any future AI proposal could affect individuals, involve personal information or create a higher risk, Derek will complete an appropriate risk assessment before proceeding and will not proceed where the risk cannot be adequately controlled.

Derek will explain material AI involvement where it would reasonably matter to a client or affected person. Clients may ask how AI has been used in work delivered by Guiding North HR. This policy does not require disclosure of routine spelling, formatting or low-risk administrative assistance where that would not affect the substance of the work.

AI prompts and outputs must not contain passwords, security credentials, confidential case details or unnecessary personal information. Working files are kept only for as long as needed and are handled in line with the Data Retention and Deletion Policy.

Questions or concerns about Guiding North HR's use of AI can be sent to derek@gnhr.co.uk. Derek will review this policy whenever an approved tool, use case or legal requirement changes and at least annually.