Introduction

Artificial Intelligence (AI) is the term given to systems or tools that can perform tasks that typically require human intelligence faster and in certain circumstances, more accurately than humans. Ways that AI can be used in the workplace include but are not limited to:

• text processing

• image generation and editing

• meeting notes and transcripts

• data analysis

• workflow automation.

Vet Sustain recognises the growing potential of AI in the workplace which, if used responsibly and appropriately in a manner consistent with the values and objectives of the organisation, can benefit Vet Sustain’s operations and the service it provides to its members and customers, enabling more effective use of staff time and supporting decision-making.

However, Vet Sustain also recognises that the use of AI can present some challenges and risks and so wishes to encourage an appropriate and responsible approach to AI in the workplace. This involves understanding its capabilities and limitations to ensure that Vet Sustain can utilise AI tools effectively while maintaining trust and integrity and complying with the law on data protection.

Purpose of the AI policy

This policy aims to provide clear guidelines and expectations on the responsible and appropriate use of AI by Vet Sustain employees and working group members, to ensure that it is used effectively and ethically while safeguarding Vet Sustain’s integrity, reputation and the privacy of its data.

Vet Sustain acknowledges that AI can only support, and not replace, professional skills, creativity and judgement.

Responsibilities

1. Managers’ responsibilities

Managers should discuss with their team members how they may be able to use AI to support their work and provide guidance and learning opportunities as appropriate.

Managers should make Vet Sustain’s AI use policy clear to all employees in their team and check that AI is being used appropriately and responsibly in their work.

Managers have responsibility for taking action should they become aware that an employee is not complying with Vet Sustain’s AI use policy by speaking to the employee concerned and flagging it with the relevant Director.

2. ’ responsibilities

Employees have responsibility for familiarising themselves with Vet Sustain’s AI policy and following the principles and guidelines below.

Principles of the AI policy

• Vet Sustain aims to develop a balanced approach to AI use in the workplace by ensuring it is properly understood, and used appropriately, responsibly and transparently.

• Employees should be mindful of Vet Sustain’s values and expected standards of behaviour when accessing, interpreting, and acting on material generated by AI in the course of their work.

• AI should only be used as a tool to aid judgement, and not to replace it. Employees should always use their professional expertise and judgement when using AI tools.

• Employees should be transparent by clearly flagging where they have used AI in their work (see Appendix 1 for more detail).

• Vet Sustain’s policies in relation to IT acceptable use, data use and data protection, including compliance with GDPR, apply when using personal AI tools. Employees should treat all data with the same level of security and care and should not place confidential Vet Sustain data into third party applications.

• In addition, employees should ensure they do not pass ownership of Vet Sustain data to organisations that own the AI tools and therefore should be extremely careful when agreeing to external terms and conditions to avoid inadvertently transferring ownership.

• Employees are encouraged to use AI tools where appropriate, sharing their experiences with AI and how they have used it successfully with their team and other Vet Sustain colleagues to help everyone understand the practical benefits and applications of AI.

• Employees who choose to use AI in their work are encouraged to maintain their AI skills and capabilities by keeping up to date with technological developments, sharing good practice, experimenting responsibly, and participating in learning activities.

‘Human-in-the-loop’ Approach:

AI can greatly improve productivity in areas like content creation, analysis, and decision-making, but it cannot replace human judgment. Without human oversight, AI may produce false information, amplify biases from its training data, and fail to understand nuance, context, or ethical implications. Human review is therefore essential to verify accuracy, maintain ethical standards, preserve trust and brand voice, and ensure accountability. The most responsible approach is a “human-in-the-loop” model, where AI provides speed and scalability while humans guide, evaluate, and make the final decisions.

Environmental Impacts of AI:

It is important that consideration is given to the growing environmental impact associated with AI and its infrastructure. Environmental impacts include; the generation of electronic waste from AI servers which often contains hazardous substances such as mercury and lead, the consumption of water for construction of AI data centres and for cooling of electronic components, the reliance on critical minerals and rare elements for the production of microchips (often which are mined unsustainably) and the huge energy requirements to power complex AI electronics. A request made through ChatGPT consumes 10 x the electricity of a similar google-based search and the true environmental impacts are yet to be quantified. Vet Sustain therefore encourages its community to be mindful of this when using AI to avoid unnecessary use.

Raising concerns

Where employees become concerned that AI is being used inappropriately or improperly in the workplace, they should raise their concerns directly with their own or the relevant line manager or Director.

All employees and volunteers are expected to comply with this AI policy and guidelines. Where an employee/volunteers actions contravene this policy, it will be managed through Vet Sustain’s disciplinary procedure and may lead to disciplinary action.

Appendix 1 - Guidelines for employees when using AI in the workplace

• You can use AI for routine tasks such as spelling and grammar checks without further checks.

• Follow the ‘humans-in-the-loop’ approach where AI provides speed and scalability while human oversight remains to guide, evaluate, and make the final decisions.

• For more complex tasks, such as analysing documents or developing policy, always critically assess the output of AI tools for accuracy and reliability and verify the results through cross-referencing with reliable sources. If in doubt, you should sense check with your line manager or a colleague.

• AI can be useful to produce a first draft or an outline of a document, but you then need to think about how it can be modified and finessed so that it is consistent with Vet Sustain’s tone of voice.

• You should never pass off content generated by AI as your own work.

• Avoid any unauthorised copying of original work by creators so as not to infringe intellectual property rights.

• Be aware that AI can have an inherent bias and perpetuate stereotypes, so when using AI, be careful that it does not produce content that is inconsistent with Vet Sustain’s EDI policy.

Large Language Models

• Examples of Large Language Models (LLMs) are online tools such as ChatGPT, Microsoft Copilot or Google Gemini.

• It is not appropriate to use Large Language Models for generating work that is outside your area of knowledge or expertise. Only use LLMs to supplement your existing knowledge and do not rely on them to produce content in unfamiliar areas.

• Thoroughly check any AI-generated content for errors, inconsistencies, and misinterpretations, especially in sensitive areas such as policy recommendations or public communications. Additionally, check for incorrect use of language and ensure that the output aligns with Vet Sustain's standards and tone of voice.

• Any documents or slides that have a significant proportion of text or images produced by AI should be clearly flagged as such. This ensures transparency and maintains the integrity of the information provided.

Images and Graphics

• AI should not be used to create images and graphics in most circumstances, particularly in order to create life-like, photo-realistic images that may not be recognisable as AI-generated and can mislead people.

• AI can only be used in this context in very limited circumstances, such as for generating small graphics such as logos or page furniture.

• AI images should not be used at all in external or network-facing communications, as they can misrepresent reality, create false impressions and remove authenticity.

Data Privacy and Security

• Ensure that the use of AI complies with GDPR and other relevant data protection regulations. Personal data should only be used in AI tools if it is anonymised.

• Vet Sustain does not currently have a secure AI search application and therefore use of AI tools will train on inputs and search queries from Vet Sustain. Therefore, personal or confidential data should never be shared in search queries. Avoid using personal accounts or unapproved applications for work-related tasks involving AI.

• Do not share sensitive or personal data with AI tools that are not secure or compliant with Vet Sustain’s data protection policies.

• Do not use any of Vet Sustain’s personal identifiable information (PII), confidential business information and member information in personal AI accounts.

Glossary of terms