Donate

Generative AI policy

Introduction

This policy sets out LandAid’s position on the use of generative artificial intelligence (AI) by its staff.

 

Generative AI offers significant opportunities to enhance our creativity, productivity, and internal efficiency, particularly in areas such as communications, fundraising, research, and grant administration.

 

As an organisation that champions innovation while being mindful of our responsibilities to young people, partners, and the environment, we are committed to using AI thoughtfully, transparently and ethically.

 

This policy is public-facing to demonstrate our approach to AI and to provide clarity on how LandAid-generated content may be produced using generative tools.

 

This policy has been informed by the UK Government Communications Service (GCS) Generative AI policy, UK Government guidance on responsible AI use, and relevant data protection guidance from the Information Commissioner’s Office (ICO). It has been adapted to reflect LandAid’s size, structure, charitable mission and risk profile.

 

Scope 

This policy applies to all staff at LandAid and covers the use of generative AI tools, including but not limited to:

  • Large language models (e.g. ChatGPT, Gemini)
  • Image generators (e.g. DALL·E, Midjourney)
  • Automated assistants (e.g. Microsoft Copilot, Grammarly)

 

This policy does not apply to non-generative tools (e.g. standard automation tools, analytics platforms) unless they include generative features.

 

Definition: commercially sensitive and confidential information

For the purposes of this policy, commercially sensitive or confidential information is any information that is not publicly available and that could reasonably cause harm to LandAid, our partners, or the young people we support if it were disclosed outside LandAid.

 

This includes, but is not limited to:

  • Board and Committee papers, including agendas, packs, minutes and action logs that are not in the public domain
  • Non-public financial information, including budgets, forecasts, reserves planning and contractual terms
  • Information relating to partnerships, negotiations or agreements that has not been publicly announced
  • Any information explicitly marked as confidential or shared with LandAid in confidence

 

Personal data (as defined under UK GDPR), including any information that could identify an individual, must not be entered into generative AI tools under any circumstances.

 

Staff may use generative AI tools to support drafting or structuring documents, including meeting notes, strategy or grant-related materials, provided that they do not include personal data or any commercially sensitive or confidential information.

 

Where using internal materials (such as meeting notes), staff must exercise judgement to ensure that:

  • Any content that is confidential, sensitive, or not intended for wider sharing is removed
  • No personal data or identifying information is included
  • The information used is appropriate for external processing

 

All staff are responsible for applying their data protection training and professional judgement when using generative AI tools. Where there is any uncertainty about whether information is appropriate to use, it should be treated as confidential and not shared.

 

Staff must not assume that generative AI tools embedded within existing software platforms provide additional data protection safeguards unless explicitly confirmed by LandAid.

 

Our Approach 

 

LandAid permits the responsible use of generative AI within the safeguards set out in this policy, where it supports our charitable mission and reflects our organisational values:

 

– Ambitious: We explore new tools that help us think bigger and work smarter.
– Responsible: We prioritise transparency, data protection and ethical decision-making.
– Tenacious: We use AI to free up time and energy for our most meaningful work.
– Inclusive: We consider the impact of AI-generated content on all audiences.
– Creative: We use AI to support innovation, not replace human ingenuity.

 

How LandAid Uses Generative AI 

 

Generative AI may be used to support:

 

  • Writing and drafting external communications content
  • Exploring ideas using custom-built personas
  • Summarising long reports or research documents
  • Ordering and structuring complex information
  • Comparing applications against internal scoring criteria
  • Automating low-value or repetitive admin tasks

 

AI will never be used to make final decisions in our grant-making or funding processes. All decisions involving funding, partnerships, or public communications remain the responsibility of LandAid staff.

 

Key Principles for Staff 

 

Do:

  • Be transparent when AI has meaningfully contributed to work.
  • Use AI to enhance your work, not replace professional judgement.
  • Keep prompts and outputs free from personal data and any commercially sensitive or confidential information (as defined in this policy).
  • Consider whether AI is necessary before using it, particularly to minimise environmental impact.
  • Review AI-generated content for bias, error, or poor-quality information.
  • Verify facts, statistics, quotations and references against reliable sources before publication.
  • Check outputs for bias, stereotyping or exclusion, particularly when referring to people with lived experience of homelessness or protected characteristics.
  • Ensure AI-generated content aligns with LandAid’s tone of voice, brand guidelines and use of UK English.

 

Don’t:

  • Use AI tools to process or analyse personally identifiable or commercially sensitive data.
  • Rely on AI for outputs that require sensitivity, safeguarding or nuance without human input.
  • Input third-party copyrighted content unless you have the right to use it, or publish outputs that replicate identifiable third-party work without appropriate checks.

 

Brand, Tone and Quality Control

LandAid’s tone of voice, use of UK English, and brand standards are central to how we communicate our mission and impact.

While generative AI tools can support drafting and idea development, they do not automatically produce content that aligns with LandAid’s tone of voice or brand guidelines.

 

Some tools used by LandAid (such as ChatGPT via an organisational account) may include additional instructions or guidance to support brand alignment. Other tools, including AI features embedded within existing software platforms (such as Microsoft Copilot), do not apply consistent organisational instructions and may produce outputs that vary in tone, style and language.

 

Staff are responsible for ensuring that all AI-generated content:

  • Is written in UK English
  • Reflects LandAid’s tone of voice and values
  • Meets the standards set out in our brand and communications guidelines

 

Where generative AI tools are used to produce public-facing content, outputs must be carefully reviewed and, where appropriate, shared with the Communications team before publication to ensure consistency, accuracy and alignment with LandAid’s brand.

 

Environmental Responsibility 

 

LandAid recognises that generative AI tools rely on energy-intensive computing infrastructure and have an environmental impact. 

 

As a responsible organisation, we expect staff to use generative AI thoughtfully and proportionately. AI should be used where it adds clear value, improves efficiency, or enhances quality – not as a default for routine tasks. 

 

Before using generative AI, staff should consider: 

 

  • Whether the task genuinely requires AI support
  • Whether the outcome could reasonably be achieved without it
  • Whether repeated or large-scale use is necessary

 

LandAid encourages efficient and purposeful use of technology in line with our commitment to responsible operations. 

 

Data Protection and GDPR Compliance 

 

LandAid is committed to complying with the UK General Data Protection Regulation (UK GDPR) and the Data Protection Act 2018 in all its activities, including the use of generative AI tools.

 

Generative AI tools must not be used to process, analyse, or store personal data unless appropriate safeguards and formal approval are in place.

 

Staff must not input personal data or identifiable information relating to young people, staff, partners, donors, grantees or other individuals into generative AI tools.

 

This applies equally to generative AI tools accessed directly and to AI features embedded within existing software platforms.

 

The use of generative AI does not replace or override LandAid’s existing Data Protection Policy, Privacy Policy, or confidentiality obligations. All staff remain responsible for ensuring that their use of AI complies with these policies.

 

Human Accountability 

 

All content and decisions generated with the support of AI remain the responsibility of LandAid staff. This is especially critical in our grant-making processes, where AI may support administration or scoring but will never replace human judgement or decision-making.

 

Transparency 

 

LandAid is committed to being open about its use of generative AI. Where generative AI plays a significant role in producing public-facing content, imagery, or digital services, we will consider whether disclosure is appropriate in order to maintain trust and transparency with our audiences.

 

We will not use generative AI in ways that could mislead audiences about the authenticity of lived experience, testimony, or personal stories.

 

Where external agencies, consultants or freelancers are engaged on LandAid work, they are expected to follow this policy when using generative AI tools and to apply equivalent data protection and confidentiality safeguards.

 

 

 

Approved Generative AI Tools

 

LandAid will maintain a register of approved generative AI tools for work-related use.

 

Staff may only use generative AI tools that have been formally approved and included on this register. The Director of Communications is responsible for reviewing and approving tools and maintaining the register.

 

When assessing whether to approve a tool, LandAid will consider:

 

  • Data protection and data handling arrangements
  • Storage and processing location of data
  • Security safeguards and confidentiality protections
  • Transparency regarding how inputs may be used or retained
  • Alignment with LandAid’s values and risk appetite

 

Staff may propose additional tools for review. No new generative AI tool may be used for LandAid work until it has been approved.

 

LandAid currently uses ChatGPT Plus via an organisational account. Staff must not upload or input personal data or commercially sensitive or confidential information into this tool, in line with this policy.

 

LandAid also recognises that some generative AI features are embedded within existing software platforms (such as Microsoft 365 applications). These tools may be used by staff for work purposes, provided they are used in accordance with this policy.

 

Where generative AI tools are available within existing platforms but are not covered by enterprise-level data protection agreements, they must be treated in the same way as external AI tools. This means that personal data and commercially sensitive or confidential information must not be entered into these tools under any circumstances.

 

Staff must not assume that generative AI tools embedded within existing software platforms provide additional data protection safeguards unless explicitly confirmed.

 

This policy applies specifically to generative AI tools. Standard automation or workflow platforms (such as Microsoft Power Automate) remain governed by LandAid’s existing IT and data protection policies.

 

 

Training and Internal Guidance 

 

Before using generative AI tools for LandAid work, staff are expected to complete internal training or guidance to ensure a consistent and responsible approach.

 

This training will help ensure that all team members:

 

  • Understand the capabilities and limitations of generative AI
  • Use AI tools in ways that align with LandAid’s values and mission
  • Are aware of ethical, environmental, and data protection considerations
  • Apply good judgment in reviewing and editing AI-generated content
  • Understand the difference between consumer, embedded and enterprise AI tools, and the implications for data protection

 

The training will be led or coordinated by the Director of Communications, and updated as tools evolve.

 

Oversight and Review 

  • The Director of Communications is responsible for oversight of this policy.
  • Staff do not need advance permission to use generative AI tools if they follow this policy.
  • This policy will be reviewed annually, or earlier if significant changes in technology, tools, or risk require it.

Questions or Concerns 

If you are unsure whether a particular use of AI is appropriate, please contact the Director of Communications. We welcome open dialogue as we continue to learn and evolve our approach to generative AI.