Guide

NHS Communications Artificial Intelligence Operating Framework

Principles, expectations and safeguards to ensure AI is used in ways that are ethical, inclusive, transparent and anchored in trust.

1 October 2025

This operating framework supports NHS communications and engagement professionals to explore and adopt artificial intelligence (AI) in a way that is ethical and effective, and promotes trust, inclusion and transparency.


It offers clear, practical guidance to empower the responsible use of AI tools and technologies, enabling NHS communications and engagement teams to build capability, develop confidence and collaborate effectively with IT, digital, information governance and clinical colleagues. This framework is designed to help professionals advocate for safe access to AI platforms and tools that can enhance patient engagement, improve efficiency and drive more personalised communication.


This is not a policy document, nor a replacement for local protocols. Instead, it offers a shared foundation that NHS communicators can adapt and apply based on local needs, in partnership with other key stakeholders within their organisation. It will be updated regularly (at least every 12 months) to reflect changes in technology, regulation and best practice.


This framework forms part of the NHS Communications AI Taskforce’s national priorities, supported by the NHS Confederation.

Foreword

Enabling responsible AI use in NHS communications

The emergence of generative artificial intelligence (AI) offers unprecedented opportunities to transform how the NHS communicates with its patients, staff and communities. From simplifying language to scaling insight generation and co-creating more inclusive campaigns, these tools have the potential to enhance both the effectiveness and reach of our communications efforts.

Yet access to these tools remains inconsistent across the NHS. Some NHS communications and engagement teams are early adopters, while others are hampered by infrastructure, uncertainty or challenges navigating governance. This variation has the potential to create a fragmented communications landscape – where the benefits of innovation are not realised across the NHS, for patients, carers and staff.

This national operating framework, which will evolve as the AI landscape does, responds directly to that challenge and follows an in-depth report published in June 2025 that described the opportunities and challenges with AI adoption in NHS communications.

It does not propose a solution to every use case, nor does it mandate a one-size-fits-all approach. Instead, it offers a structured foundation – an enabling framework that NHS communications and engagement professionals can discuss with their teams and colleagues, including chief information officers, information governance (IG) leads and executive sponsors, to begin – or strengthen – local dialogue. Its purpose is to unlock proportionate access, foster partnership working and support the responsible exploration of AI use in NHS communications.

Together, we will shape a future where AI enhances – not replaces – the human connection at the heart of NHS communication

This framework sets out the principles, expectations and safeguards needed to ensure AI is used in ways that are ethical, inclusive, transparent and anchored in trust. As national case studies continue to evolve, this living document will hold that space for iteration.

We invite communications and engagement teams, digital leaders and IG professionals to use this framework not just as a governance tool, but as a catalyst for collaboration. Together, we will shape a future where AI enhances – not replaces – the human connection at the heart of NHS communication.

NHS Communications AI Taskforce

Introduction, purpose and scope

AI is already beginning to reshape how the NHS delivers services – from diagnostics and appointment scheduling to staff engagement. Communications and engagement teams are increasingly using AI to simplify technical language,  generate first drafts of content, analyse complex information and improve accessibility.

According to the NHS Communications AI Taskforce’s June 2025 report, over half of NHS communications professionals are already using AI tools in some capacity, often informally, to work faster and with greater confidence. These early use cases highlight AI’s potential as a creative assistant, supporting – not replacing – human expertise, and underscore the need for a structured, supported approach throughout the NHS.

This framework offers a clear, trusted point of reference for communications professionals to engage with AI

This framework exists to support that momentum – by offering a clear, trusted point of reference for communications professionals to engage with AI confidently and collaboratively through the work of the NHS Communications AI programme. Its purpose is to:

  • Enable NHS communicators to proactively engage with AI opportunities in their day-to-day work.
  • Provide shared language and principles to support dialogue with key internal stakeholders, including information governance, digital and IT colleagues.
  • Build trust in responsible AI use by embedding NHS values into local practice and decision-making.

This framework is not a replacement for local policy. Instead, it serves as a
national scaffold to help communications professionals shape safe, equitable
and effective AI adoption in a way that reflects the needs of their patients,
staff and communities.

We encourage all NHS organisations to apply this framework in partnership
with their key internal stakeholders such as IG and digital teams – ensuring
that communications and engagement staff have appropriate access to
tools, support to use them responsibly, and the freedom to explore where AI
can make the greatest difference.

Definition of terms

This section introduces key terms that support a shared understanding of AI in communications contexts. These definitions reflect how concepts like generative AI, machine learning and natural language processing are already beginning to shape the way NHS communicators work. They are not intended to be exhaustive or overly technical, but to equip professionals with practical language when engaging with colleagues in digital, data and governance functions.

Artificial intelligence (AI): computer systems that can perform tasks typically requiring human intelligence – such as recognising images, generating speech, understanding language or making predictions.

Why it matters: AI underpins many of the emerging tools used in content generation, chatbots and data analysis.

Generative AI: A type of AI that creates new content – like text, images, audio, or video – based on patterns it’s learned. For example, generating tailored patient letters or creating imagery for digital campaigns.

Why it matters: Generative AI opens up creative possibilities for communications, but requires oversight to ensure accuracy, guardrails against bias and accessibility.

Data protection impact assessments (DPIAs): A formal assessment to identify risks when using personal data in a new way or system. Often required for AI initiatives.

Why it matters: Communications teams may need a DPIA when developing tools that handle audience feedback, generate content from datasets, or use internal information to train AI systems.

Machine learning: An approach within AI where systems learn from data to improve over time without being explicitly programmed.

Why it matters: Even if not visible to end users, machine learning powers tools that help analyse feedback or optimise audience targeting.

Natural language processing (NLP): A branch of AI that helps computers understand and generate human language – spoken or written.

Why it matters: Many communications tools use NLP to summarise complex content, answer queries, or rewrite language for clarity.

Data processing: Any activity involving data – collecting, storing, analysing, adapting, sharing, or deleting it.

Why it matters: Many AI tools process data behind the scenes; understanding this term helps identify when IG consultation is needed.

Robotic process automation (RPA): Software that mimics repetitive administrative tasks, such as extracting data or populating forms. It doesn’t involve learning or generating new content.

Why it matters: While not AI in a generative sense, it can still support back-end comms workflows –like formatting reports or moving files.

Responsible AI principles for communications

When using AI in communications, NHS teams should adopt principles that support innovation and accountability. These aren’t just technical standards – they reflect our shared values of trust, inclusivity, and transparency in public service. Each principle below helps ensure that AI-enabled communication strengthens – not compromises – our connection with the people we serve:

Fairness 

AI tools should support equitable communication. For example, image generators must reflect the diversity of the NHS population across race, age, disability, gender and lived experience. AI-generated copy should be reviewed for accuracy, bias or stereotyping. 

Reliability and safety 

AI systems should work as expected, across contexts and audiences. Communications and engagement teams should regularly assess whether tools generate appropriate, relevant and safe outputs – especially when used in patient-facing materials.

Privacy and security 

AI must be used in partnership with digital and information governance (IG) teams to uphold confidentiality and data security. NHS staff should never input personal or sensitive information into public AI tools (such as ChatGPT) and should follow local guidance on platforms approved by the organisation.

Inclusiveness 

AI should help make communications more accessible, not less. This means using tools that support plain language, readability and inclusive design – especially for individuals with diverse language backgrounds, literacy levels or accessibility needs. 

Transparency*

NHS staff should be clear when content is generated or supported by AI, especially for public / patient facing content. Transparency builds trust – especially if content is later challenged or shared widely. AI can assist in writing, but humans remain responsible for its message. 

Accountability 

People – not machines – must be accountable for communications outputs. AI tools should be used to augment professional judgement, not replace it. Every output should be reviewed and sense-checked before use.

*Some NHS organisations use a blanket statement on their website rather on every piece of communication. 

For example: 

Artificial intelligence in Central and North West London NHS Foundation Trust (CNWL) 

At CNWL, we use artificial intelligence (AI) to assist in processing information and creating content. This means that advanced computer programmes help us in writing and organising data. Our aim is to ensure that the information we provide is accurate, helpful, and easy to understand. Using AI enables us to work more efficiently and more productively. We thoroughly check and review everything to ensure it meets our high standards. We have protocols in place to ensure that staff know how to use AI safely. We believe AI is a valuable tool that helps us provide you with high-quality healthcare. If you have any questions about our use of AI, please feel free to ask. We're always here to help

Duties, accountabilities and responsibilities for NHS communicators

NHS communicators have a pivotal role to play in shaping how AI is introduced, used and trusted within their organisations. The responsibilities below are grouped thematically to reflect this leadership opportunity – placing strategic partnership and professional judgement at the heart of safe, effective AI use.

Strategic leadership and advocacy

Directors of communications and engagement, and senior leads, are encouraged to: 

  • Champion the responsible adoption of AI within communications, building organisational understanding and buy-in across staff and patient/ carer groups.
  • Collaborate with chief information officers, chief clinical information officers and information governance leads to ensure communications teams have appropriate, safe access to AI tools.
  • Promote a culture of curiosity, experimentation and accountability in the use of AI for communication, using this framework as a guide.
  • Ensure communications staff receive support and training in using AI responsibly.

Everyday use: professional standards in practice

All communications professionals using AI tools should: 

  • Apply professional judgement to verify the accuracy, tone and inclusivity of any AI-generated content.
  • Flag potentially inappropriate, biased or inaccurate outputs and avoid using generative AI tools for decision-making without human oversight.
  • Use AI only on corporate devices, and do not enter any personal, confidential or sensitive data into public platforms (such as ChatGPT).
  • Sense-check images created using AI to ensure fair representation and alignment with NHS values.
  • Ensure any outputs comply with copyright and intellectual property guidance.
  • Consider that information generated by AI may fall within scope of information access requests and require explanation or justification.
  • Ensure any mandatory training, especially information governance, is up to date.

Responsible collaboration with digital, information governance and IT

Communications teams should: 

  • Work closely with key internal stakeholders – including digital, IT and IG colleagues – to co-design safe ways of introducing AI tools.
  • Notify IG teams if publicly available AI tools (such as ChatGPT, Gemini or Microsoft Copilot) are being used as part of work tasks to ensure compliance with organisational policies.
  • Engage with the data protection impact assessment (DPIA) process for any initiatives involving personal data or where there is a potential privacy impact – even if only for testing or training purposes.
  • Treat any data that could indirectly identify individuals (such as rare conditions or local demographics) with the same care as directly identifiable information.
  • Proactively raise any concerns about AI (tools) behaviour or risks through organisational reporting channels.

Framework document requirements for communications

This section outlines the processes and considerations that enable communications teams to adopt AI tools safely, transparently and in line with national guidance. These are not barriers – they are enablers of trust, collaboration and consistent access across the NHS.

Defining purpose, outcomes and values

Start with clarity. When proposing or using AI tools, define: 

  • What problem the tool will help solve
  • What communications value it adds (such as time saved, better tone, more accessible language)
  • Whether data is involved, and if it is, whether a lawful basis for processing is needed
  • Work with your IG colleagues to confirm whether anonymised data is sufficient or whether a DPIA is required. When in doubt, ask – co-designing risk mitigation early will speed up safe access later.

Using freely available AI tools (for example ChatGPT, CoPilot, Gemini)

AI Assistant Apps - ChatGPT, DeepSeek, Anthropic Claude, Perplexity, Google Gemini, Microsoft Copilot - on smartphone

Many communications professionals are already using LLM-based tools for creative support with writing, brainstorming or accessibility checks. When doing so: 

  • Never use real patient, staff or business-sensitive data.
  • Do not enter queries that you wouldn’t want to become public, for example subject to FOI.
  • Sense-check outputs for tone, bias and accuracy.
  • Inform your IG team if using these tools regularly.
  • Treat these tools like other creative references, not definitive sources.
  • Remember that accountability remains with the human – not AI.

Procuring and piloting AI-enabled communications tools

When adopting or requesting new tools:

Responsible transparency and engagement

All AI use in communications should be transparent to staff and (where relevant) to the public. Teams should: 

  • Make it clear when AI supports content generation (note the CNWL example).
  • Be prepared to explain how outputs were created or reviewed.
  • Build inclusive practices into content creation – from diverse training data to representative imagery.
  • Ensure local communications about AI use are accessible, plain-English and aligned with NHS values.

Consultation

This framework was shaped through collaboration with communications professionals, digital leaders and information governance specialists across the NHS, coordinated through the NHS Communications AI Taskforce. 

It reflects broad input from across the NHS and draws on practical insights captured through the taskforce’s national listening sessions, working groups and June 2025 sector-wide report. 

This is not intended to be a perfect document. We will refine it further, taking into consideration feedback and changes in the AI landscape, both technical and legal. We encourage organisations to treat this framework as a living document and reference point – one that evolves through continued dialogue, shared practice and feedback from the professionals who use it.

Training

To ensure the responsible and effective use of AI in NHS communications, staff must be equipped not only with technical understanding, but with the professional judgement to apply AI in ways that align with NHS values and local context. 

Training should not be viewed as a one-off requirement, but as a core component of long-term professional development

The NHS Confederation, supported by the NHS Communications AI Taskforce, will coordinate national training efforts. This will include workshops, seminars and toolkits designed to support: 

  • Awareness of AI’s opportunities and limitations in communications
  • Responsible use of generative AI tools, including risk mitigation and ethical safeguards
  • Real-world use cases shared by peers across the NHS
  • Guidance on working effectively with IG, digital and procurement colleagues
  • Hands-on confidence in reviewing, adapting and sense-checking AI-generated content. 

Training should not be viewed as a one-off requirement, but as a core component of long-term professional development. We encourage organisations to share learning, feedback and resources that can support a thriving, connected communications community as AI capabilities continue to evolve over time.

Monitoring compliance

As this framework is guidance, not prescriptive policy, local implementation must be shaped by individual organisations and led by their senior communication and engagement professionals. 

Directors of communications and engagement are encouraged to ensure the principles in this framework are actively embedded through: 

  • Regular conversations about AI use within teams
  • Inclusion of AI in team development plans, use-case reviews and knowledge sharing
  • Proportionate documentation – such as checklists for tool use or DPIA triggers
  • Informal audits or spot checks where appropriate, in collaboration with digital and IG colleagues.
  • Ensuring that AI learning and development is part of the annual objectives. 

The NHS Communications AI Taskforce will maintain national oversight of progress and uptake. Its role is to convene learning, highlight emerging good practice, and adapt the framework based on system-wide insight. It will not monitor individual organisations or enforce local compliance. 

Monitoring is not about constraint – it is how we ensure our collective AI journey is transparent, safe, and grounded in the principles set out in this framework.

Arrangements for review

This framework will be reviewed and updated at least every 12 months by the NHS Communications AI Taskforce. Earlier revisions may be triggered by exceptional circumstances, emerging technologies, organisational change or relevant shift in legislation or national guidance. 

We encourage NHS organisations to treat this as a living framework – one that can be adopted and iterated locally, in partnership with IG, digital and communications teams. 

Feedback, insights and case studies from across the NHS will inform each update. As new tools, risks, and opportunities emerge, we will ensure the framework reflects the needs and realities of the professionals it supports. 

To support this dynamic approach, NHS communicators are encouraged to stay engaged with ongoing learning, share what works in practice, and help shape the next evolution of responsible, values-driven AI use.

Dissemination

The framework will be made available through the NHS Confederation website and shared through regional NHS communication channels to ensure broad consistent access across the system. 

Its publication marks the start – not the end – of a wider conversation. We encourage NHS communications and engagement teams, IG colleagues and digital leaders to use this document as a starting point for local discussions about AI. 

Widespread dissemination will help ensure equity of access to insight and opportunity, reduce duplication of effort and build a shared baseline for responsible innovation. As local use cases emerge, this framework will evolve in parallel, rooted in collective learning and NHS wide participation

Equality and inclusion

AI systems used in NHS communications should reflect the diversity and lived experiences of the people we serve. This includes the language we use, the images we create, and the insights we draw from data. 

Inclusive communication is a core NHS value. AI should strengthen, not undermine, that commitment

Communications and engagement professionals should proactively consider how their use of AI can either reduce or unintentionally reinforce existing inequalities. This includes:

  • Ensuring diverse representation in AI-generated visuals and written outputs
  • Reviewing how training data or tool design may embed bias
  • Actively involving underserved audiences in shaping content or tool evaluation where feasible. 

Inclusive communication is a core NHS value. AI should strengthen – not undermine – that commitment. Equality impact assessments should be carried out in line with organisational policies.

General Data Protection Regulation (GDPR)

AI systems that process personal or sensitive data must comply with the UK General Data Protection Regulation (GDPR) and the Data Protection Act 2018. This includes: 

  • Completing a data protection impact assessment (DPIA) where processing is likely to create high risk.
  • Minimising the use of identifiable data wherever possible, and anonymising data before training or testing AI models.
  • Consulting with your organisation’s information governance team at the earliest opportunity. 

Data privacy is not just a legal duty – it is key to public trust. When used well, AI can improve communication and insight generation without compromising confidentiality or ethical standards.

About the NHS Communications AI Taskforce

This framework forms part of a wider project being led by the NHS Communications AI Taskforce in partnership with the NHS Confederation. The taskforce provides strategic leadership and oversight for all the workstreams that are outlined in this report.