Key points
This report forms part of a wider project being led by the NHS Communications AI Taskforce in partnership with the NHS Confederation. It presents the findings of a research study that builds on the foundations set out in an engagement paper on AI in NHS communications, published in December 2024.
The primary aim of this latest research was to understand how NHS communications professionals are using AI tools; where there is potential to apply these tools more effectively; what barriers, risks and capability gaps need to be addressed; what kinds of support, training and governance will enable safe, ethical and confident AI adoption.
Our key findings are:AI is beginning to reshape how communications teams across the NHS work. These technologies offer new ways to support content creation, improve accessibility, analyse feedback and streamline everyday tasks.
Adoption of AI is uneven. More than half of respondents (55 per cent) to our survey reported using AI tools in their role, with a further 41 per cent interested in starting to use them. However, access to tools and skills is uneven and much use remains informal or experimental. Much of the profession describes itself as being at ‘beginner’ status.
AI is seen as an assistant, not a replacement. Across our survey and focus groups we ran, communications leads described AI as a useful support tool that helps draft content, simplifies complex information and speeds up repetitive tasks. Human oversight remains essential for accuracy, empathy and alignment with NHS values.
Efficiency and confidence gains are clear. AI is helping the teams using it to work faster and with greater confidence, particularly when simplifying technical language into plain English, generating first drafts or analysing feedback from patients and staff.
Governance and trust matter. Informal use of AI tools, particularly generative tools such as OpenAI ChatGPT, is widespread but often outside formal approval processes. However, we did not find any evidence of communications professionals using it to produce clinical or other patient-related information without the appropriate checks by professionals. There is a clear need for governance, guidance and risk management to ensure safe, responsible use. An issue is how and when the use of AI should be transparently declared, especially in the production of patient-facing information that is medical in nature.
AI has the potential to level the playing field. By supporting less experienced staff and smaller teams, AI can help ensure that NHS communications remain high quality and accessible for staff and patients - even in resource-constrained environments.
Training and capability building are urgent needs. Both senior leaders and delivery teams expressed a need for role-specific training. Directors want to understand the strategic, ethical and governance implications, while operational teams need practical skills for effective use.
To take this forward, the NHS Confederation and the taskforce have identified five strategic priorities for action, each with a clear delivery route:
1. Develop a national operating framework for the use of AI in NHS communications.
2. Establish an NHS Communications AI Network. This will connect peers to develop their use of AI by sharing practical insights and feedback on approaches that will help NHS organisations use AI to improve how they communicate with staff, patients and communities. This is launching on 26 June.
3. Develop an ethics framework, setting out the principles and values to guide our use of AI and address crucial issues such as data protection, privacy, consent, fairness, transparency and human oversight.
4. Establish an online innovation and training hub as a centre of excellence platform for collaboration and knowledge exchange.
5. Develop a long-term monitoring and evaluation system to ensure accountability and continuous improvement.
How NHS communications professionals are adopting AI tools and what is needed to enable safe, ethical and confident use.
Background
AI is beginning to reshape how communications teams across the NHS work. These technologies offer new ways to support content creation, improve accessibility, analyse feedback and streamline everyday tasks. This is happening at a time when NHS communications teams face significant resource pressures and rising demand. This is also combined with the need to engage with diverse patient groups and other stakeholders with clarity, empathy and trust.
As AI capabilities evolve rapidly, NHS communicators are increasingly curious about how these tools could help them meet these challenges. However, access to tools, confidence in their use and organisational readiness remain highly variable across the health service. Without clear governance, training and shared learning, there is a risk that early AI adoption could become fragmented, inconsistent or widen capability gaps between teams.
At the same time, with the right support and safeguards in place, AI has the potential to help level the playing field, enabling smaller teams, less experienced practitioners and stretched communications functions to produce high-quality, audience-focused work more efficiently.
“It’s already making a difference, but who gets to benefit depends on who has the tools, the skills and the permission to use them.”
Focus group participant.
About this research
This report forms part of a wider project being led by the NHS Communications AI Taskforce in partnership with the NHS Confederation. It presents the findings of a research study, consisting of a major survey, focus groups and desk research, that builds on the foundations set out in an engagement paper on AI in NHS communications, which was published in December 2024 by the NHS Communications AI Taskforce in partnership with the NHS Confederation.
The engagement paper set out a shared vision and guiding principles. It also outlined proposed action areas for the use of AI across NHS communications, including a draft ethical framework. It was informed by nine months of engagement with communications teams working across the NHS London region and input from independent industry experts.
The actions and initiatives set out in the original engagement are outlined in Appendix 1.
From December 2024 to February 2025, following the publication of the engagement paper, views were gathered from communications professionals, industry partners and other stakeholders to answer key questions about where AI was already being used, where there was ambition to do more and what risks and barriers needed to be addressed.
This research represents the next step in that process. It aims to move from initial regional consultation to a deeper evidence base on how AI is already being used within NHS communications, where barriers remain, and what support is needed for safe, effective and ethical adoption.
This research underpins the development of several activities that the NHS Communications AI Taskforce has been prioritising:
- Developing a national operating framework for the use of AI in NHS communications.
- Establishing an NHS Communications AI Network, which will connect peers to develop their use of AI by sharing practical insights and feedback on approaches that will help NHS organisations use AI to improve how they communicate with staff, patients and communities. This is being launched on 26 June.
- Developing an ethics framework, setting out the principles and values to guide our use of AI and address crucial issues such as data protection, privacy, consent, fairness, transparency and human oversight. (A first draft was also set out in the engagement paper and is include as Appendix 2 in this report.)
- Establishing an online innovation and training hub as a centre of excellence platform for collaboration and knowledge exchange. Our latest report also identifies a requirement for training on two levels for senior leaders and operational communications roles.
- Developing a long-term monitoring and evaluation system to ensure accountability and continuous improvement.
Together, the engagement paper and this study provide the evidence base for developing a national AI operating framework for NHS communications, alongside supporting ethics guidance, governance recommendations and a programme of shared learning.
The operating framework, which we will produce in summer 2025, is intended to be guidance and we will support NHS communications professionals to implement it through our NHS Communications AI Network, working groups and training programmes. However, it will be for local NHS organisations to then determine how this guidance informs their formal AI communications policies and approach in line with their governance arrangements.
Purpose of the research
The primary aim of our latest research was to understand:
- how NHS communications professionals are using AI tools
- where there is potential to apply these tools more effectively
- what barriers, risks and capability gaps need to be addressed
- what kinds of support, training and governance will enable safe, ethical and confident AI adoption.
The research was designed to listen carefully to the real-world experiences of NHS communicators - spanning senior leaders, managers and operational teams - and to reflect the diversity of organisations and roles across the sector.
How the research was carried out
The research was delivered through key stages:
| Stage | Activity | Purpose |
| Survey | National survey (September 2024 – March 2025) with more than 414 NHS communications professionals responding. | Captured patterns of current AI use, confidence levels, risks and training needs. |
| Focus groups | Three national focus group sessions held in March 2025, engaging a cross-section of NHS communicators from different regions, roles and organisation types. | Explored key themes in more depth, including aspirations, risks, governance issues and capability gaps. |
| Thematic analysis | Integrated analysis of quantitative and qualitative data, with validation from the AI in NHS Communications Taskforce and sector experts. | Identified shared patterns, emerging priorities and recommendations for action. |
This combination of approaches provides a robust evidence base for the recommendations in this report, rooted in the day-to-day realities, ambitions and concerns of NHS communications professionals.
What this report provides
This report presents the findings from the research and sets out practical recommendations for enabling the safe, effective and innovative use of AI in NHS communications. It aims to:
- support NHS communicators at all levels to navigate the fast-evolving AI landscape with confidence
- provide the evidence base for the planned operating framework and associated ethics and governance guidance
- inform the development of the NHS Communications AI Network, designed to promote peer learning, resource sharing and innovation across the sector.
Above all, the report reflects the core principle established in the engagement paper: that AI should enhance, not replace, human communications, helping NHS teams to maintain the trust, empathy and patient focus that sit at the heart of effective engagement.
How NHS communications teams are using AI tools
AI tools are already being used across many NHS communications teams, but take-up is uneven and a large proportion of teams have yet to grasp the opportunities. Where AI is being used, there is a wide variety of use cases emerging - from content drafting and social media adaptation to meeting summarisation and feedback analysis. However, this use is often informal, experimental and shaped by what individual teams or practitioners can access, rather than by organisation-wide strategies or consistent governance.
This section outlines the current patterns of AI adoption in NHS communications, the tools being used, how confident staff feel using them and where these tools are making a difference.
Strong interest, but uneven adoption
The national survey confirmed a high level of curiosity and enthusiasm about AI across NHS communications professionals:
This points to significant momentum and appetite for experimentation, even where formal policies or organisational strategies are not yet in place.
However, much of this current use falls into what participants described as ‘informal adoption’, with individuals or small teams exploring tools like OpenAI ChatGPT or Microsoft Copilot outside formal approval processes. Access to tools is often dependent on personal initiative, trial accounts or discretionary budgets, leading to variation in use both within and between organisations.
“There’s lots of interest, but it feels like we’re in the wild west - we’re all testing things in our own way.”
Focus group participant
This patchwork approach risks widening capability gaps between those with access and those without.
The most commonly used AI tools
Survey participants reported using a range of AI tools, often in combination:
While some teams are beginning to access enterprise-level tools through organisational licenses, much use remains at the pilot or individual level. The focus groups highlighted that even where licences exist, uptake can be limited by gaps in prompting skills, training and confidence.
It is worth noting that GCS Assist is a tool developed by the Government Communications Service for communications professionals working in government departments. Communications professionals working in frontline NHS organisations are currently not allowed to access it, but some professionals working in national NHS organisations, such as NHS England, have been able to access it on a limited basis (hence why the tool was cited in our survey). During the course of our research, GCS announced that it would be making the source code for GCS Assist openly available. We are exploring how to use this to create an NHS communications equivalent tool for frontline teams.
AI confidence levels: wide variation across the sector
The research revealed a mixed picture when it comes to confidence in using AI effectively:
This highlights the urgent need for capability-building support, particularly for those just beginning to explore these tools. Many participants reported learning through trial and error, informal peer networks or online resources, with little access to structured training.
“I know I’m not using it to its full potential yet, but even at a basic level it’s helping me work faster and smarter.”
Focus group participant
Early impact: efficiency gains and creative support
The survey findings suggest that NHS communications teams already see practical benefits from AI use, particularly in boosting efficiency, improving confidence and supporting creativity. The most cited use cases of AI in communications in the survey are outlined in the chart below:
However, participants often noted that AI does not always reduce total time spent on tasks. Instead, it changes where time is spent, freeing up capacity for higher-value activities such as strategic thinking, creative refinement or deeper audience engagement with patient, staff or community groups.
“It’s not about cutting corners. It’s about giving us back some headspace to do the things that require human skills.”
Focus group participant
Summary: momentum for adoption, but support is needed
The research paints a picture of strong enthusiasm and early experimentation, but also highlights key gaps that could hold back safe, effective adoption:
- Adoption is starting to gain momentum, but large proportions of the NHS communications profession are not using AI.
- Access to approved tools remains uneven.
- Confidence and skills in using AI tools vary widely.
- Shadow use continues in the absence of clear governance and guidance.
- Shared learning across organisations is limited.
These insights underscore the need for both guardrails and green lights, combining clear governance with the permission and support to innovate safely.
The next section explores the specific areas where AI is being applied and where NHS communicators see the greatest opportunities for impact.
Where AI is being used in NHS communications
The research identified a broad and growing range of ways that NHS communications teams are applying AI tools to support their work. While many teams are still at the early stages of adoption, the evidence shows that AI is already helping to improve efficiency, unlock creativity and support insight generation, particularly in the context of limited resources coupled with rising demand.
These emerging use cases demonstrate how AI can enhance, rather than replace, human communication skills. Across the sector, communicators are finding that AI can help them get started faster, simplify complex tasks and focus their time on higher-value work.
“It’s not about replacing people, it’s about giving us the tools to do more of the good work we want to do, with the time and headspace we’ve got.”
Focus group participant
Key application areas for AI in NHS communications
The research identified eight core areas where AI tools are already being applied, alongside opportunities for future development.
| Application area | How AI Is currently supporting communications work |
| Administrative support | Transcribing and summarising meetings or engagement events; capturing key actions, decisions and themes. |
| Drafting and editing content | Producing first drafts of press releases, staff briefings, newsletters, blogs and leadership messages, with human review and editing to ensure accuracy and empathy. |
| Social media adaptation | Tailoring content across platforms (LinkedIn, Facebook, Instagram); adjusting language and tone for different patients of other stakeholder groups. |
| Simplifying language and improving accessibility | Translating technical or clinical information into plain English. Enhancing readability for diverse audiences, including people with dyslexia or ADHD. |
| Strategic planning and idea generation | Brainstorming campaign ideas, headlines and messaging options. Supporting content audits and identifying gaps or duplication. Sourcing templates and developing and testing draft communications plans. |
| Custom GPTs and prompt libraries | Developing shared prompt templates and custom AI models that reflect NHS tone of voice, brand guidelines and inclusivity principles. |
| Data analysis and trend spotting | Exploring sentiment analysis and thematic insight generation from larger datasets (eg. staff feedback, public consultations, patient engagement). |
“AI gets you 70 per cent of the way - but it still needs a human touch to get it right.”
Focus group participant
Administrative support: reducing the burden of routine tasks
Many teams reported using AI to support routine administration, particularly in meeting management, including:
- transcribing engagement events and workshops
- drafting minutes and action logs
- summarising discussions into key points and next steps.
Tools like Microsoft Copilot and automated transcription services were highlighted as helpful in freeing up time for more strategic work.
“It saves time on admin and helps us focus on what matters: the engagement and the actions.”
Focus group participant
Drafting and editing content: faster first drafts, not finished products
The most common AI use case reported was drafting written content. Communicators are using tools like OpenAI ChatGPT, Microsoft Copilot and GCS Assist to generate:
- initial drafts of press releases, blogs, staff updates and chief executive messages
- structured content that can be refined and tailored by human communicators.
Participants were clear that AI can help accelerate the drafting process but should always be accompanied by human review for tone, empathy and factual accuracy. We did not find any evidence of communications professionals using it to produce clinical or other patient-related information without the appropriate checks by professionals.
“AI helped us with a tricky piece of stakeholder communication. We had to get the tone right.”
Focus group participant
Social media adaptation: tailoring messages for platforms and audiences
AI tools are supporting NHS teams to:
- adapt core messages for different social media channels
- adjust length, tone and style to meet each platform's expectations
- avoid repetition across posts while maintaining consistent messaging.
This has enabled faster content turnaround and improved audience targeting.
“We can now produce content that’s much more audience-specific, instead of just copy-pasting the same text everywhere.”
Focus group participant
Simplifying language and improving accessibility
Ensuring that communications are clear, inclusive and easy to understand is a top priority for NHS teams. AI is helping to:
- simplify jargon-heavy or technical language into plain English
- improve readability for people with specific needs, including neurodivergent audiences
- support the creation of more accessible materials for patients, staff and the public.
“It cut the content in half, removed repetition and gave us something we could actually publish.”
Focus group participant
Survey and feedback analysis: faster insight generation
Many teams described using AI to help analyse and summarise qualitative data from:
- staff and patient surveys
- consultation responses
- internal engagement exercises.
Tools such as Microsoft Copilot are helping teams make sense of open-text responses more quickly, identifying themes and key points for action.
“It’s not just about speeding things up, it’s about making sense of the feedback so we can act on it.”
Focus group participant
Strategic planning and idea generation: AI as a creative partner
AI is increasingly being used as a tool for creative support and strategic thinking, including:
- brainstorming ideas for campaign themes, messaging, or headlines
- supporting content audits and identifying where duplication or gaps exist
- suggesting alternative framing or ways to approach complex issues
- developing and testing draft communications plans.
Smaller teams or individuals who may have fewer colleagues to bounce ideas off particularly valued this.
“It’s like having another strategist in the room, just one who never gets tired.”
Focus group participant
Custom GPTs and prompt libraries: supporting consistency and quality
NHS organisations are beginning to explore the development of:
- custom GPT models trained on NHS brand guidelines, tone of voice and inclusivity standards
- shared prompt libraries to help less experienced users get effective results without needing advanced prompting skills.
These approaches are helping to improve quality, consistency and confidence in how AI tools are used across teams.
“We’ve created a custom GPT that understands our audience and even our inclusivity guidance. It asks questions before generating content so staff get what they need, even if they’re new to prompting.”
Focus group participant
The research shows that AI is already supporting a wide range of NHS communications tasks, with benefits in both efficiency and creativity. But adoption remains uneven, shaped by variation in tool access, confidence and organisational culture.
These findings underline the need for shared resources, clear governance and capability-building support to help teams use AI safely, consistently and effectively across the sector.
Risks, barriers and challenges
While interest in using AI across NHS communications is high, the research highlights several risks, barriers and cultural challenges that could prevent teams from realising its benefits fully and safely. These challenges are not just technical, they relate to governance, capability, trust and confidence. Without the right support, there is a risk that early AI adoption could become fragmented, inconsistent, or lead to unintended harm.
Participants were clear that the issue is not whether to use AI, but how to use it well - with the right guardrails, training and leadership support in place. Unlike the clarity provided by the Government Communications Service for the use of AI by government communications teams – backed up by training and support – frontline NHS communications teams are operating in a vacuum. This gap needs to be filled with clearer guidance and training, which is what the NHS Communications AI Taskforce and NHS Confederation are aiming to do in this space.
Much of the current use of AI in NHS communications is happening outside formal policy frameworks. This informal adoption includes individual staff or teams using free or personal accounts for tools such as OpenAI ChatGPT, often without organisational approval or oversight.
This reflects both enthusiasm and frustration: the pace of technology has outstripped the development of governance structures, leaving many staff uncertain about what is allowed, safe, or recommended.
“We’re using it because it helps - but we’re also waiting for permission.”
Focus group participant
Participants consistently called for clear, enabling governance that gives teams the confidence to use AI responsibly without fear of breaching policy or trust.
Barriers to adoption: cultural, practical and structural challenges
The research identified three types of barriers affecting the adoption and effective use of AI in NHS communications:
Barrier type | Description |
| Cultural barriers | Fear of job displacement, resistance to change, or scepticism about the value of AI. Concerns that AI could replace, rather than support, human expertise. This has been exacerbated by the latest round of budgets cuts to NHS organisations that are likely to affect NHS communications and engagement teams. |
| Practical barriers | Uneven access to AI tools due to licensing restrictions or funding gaps. Limited skills in prompting and workflow integration. Understanding the relationship between AI tools and existing data security and IT policy such as FOI requests and environmental concerns. |
| Structural barriers | Variability in digital maturity, governance and leadership engagement across NHS organisations. Lack of role-specific training or formalised support. |
These barriers often reinforce each other. For example, a lack of organisational policy can increase fear and confusion, while low confidence in prompting skills can prevent teams from making the most of even the basic functionality of AI tools.
Trust and ethics
Trust, both internally among NHS staff and externally with the public, is foundational to NHS communications. Across all focus group sessions there was a strong and consistent emphasis on the ethical deployment of AI and the critical need to preserve trust at every level. The following risks were identified and will be central to the AI operating and ethics frameworks that we will develop over summer 2025:
- Generative AI tools can produce authoritative sounding, but incorrect or misleading content. NHS communicators flagged this as a major concern, particularly where outputs may be assumed to be evidence-based without thorough verification.
- AI lacks the emotional intelligence and contextual judgment needed for sensitive or compassionate communication. Several participants warned that reliance on AI in high-emotion contexts, such as bereavement, complaints or crisis response, could damage relationships and public confidence.
- Hidden or unacknowledged use of AI in public-facing content may undermine confidence in NHS messages. Transparency about where and how AI tools are used was seen as essential to maintaining trust, particularly during times of system pressure or political scrutiny.
- Concerns were raised about AI models trained on unrepresentative or biased data, which could perpetuate inequalities in healthcare communication, particularly related to EDI contexts. Participants noted risks around messaging to marginalised communities and stressed the need for diverse testing, oversight and inclusive design.
“It’s not about stopping people from using AI, it’s about knowing where the guardrails are.”
Focus group participant
Understanding the role of AI in Freedom of Information (FOI) contexts
Some participants raised emerging questions about how AI-generated or AI-assisted content intersects with Freedom of Information (FOI) obligations. While AI tools may streamline internal communications or analysis, it remains unclear how content created or influenced by AI (especially generative models) should be recorded, stored or disclosed under FOI legislation.
- There is a need for clear guidance on AI outputs and FOI transparency, especially where AI is used to summarise feedback, generate reports or respond to public or media queries.
- NHS organisations may need to adapt records management policies to ensure AI-influenced documents are traceable and attributable.
- Training on this issue should be part of AI governance rollouts, especially for communications and FOI teams.
Environmental impact of AI use
Some focus group participants raised the issue of the environmental footprint of large-scale AI systems, particularly generative AI models. Although this is currently a lesser-known aspect, a few participants suggested that:
- AI use in the NHS should align with existing NHS Net Zero commitments
- where possible, communications teams should choose low-energy AI applications, avoid overuse, and promote digital sustainability
- environmental impacts of AI should be monitored alongside ethical risks, particularly as usage scales.
This will be explored further as part of the development of the national operating framework.
The skills gap: a barrier to confident, safe use
Across both the survey and focus groups, there was a strong consensus that training and capability building are urgent priorities. Many NHS communicators reported learning about AI through informal routes, such as self-directed online research, trial and error, or word of mouth.
This creates a risk that skills and confidence will remain uneven across teams, with some early adopters racing ahead while others are left unsure how to start.
The research identified two distinct training needs:
- For senior leaders and directors: understanding ethical considerations, governance responsibilities, risks and strategic implications.
- For operational communications teams: developing practical prompting skills, integrating AI safely into workflows and understanding where human oversight is critical.
“I’m excited to learn, but right now it feels like I’m making it up as I go along.”
Focus group participant
Balancing innovation with safe practice
Participants were clear that the challenge is not simply about limiting or controlling AI use, it is about creating the conditions for safe experimentation and responsible innovation. Communicators emphasised the need for:
- a clear policy framework that supports innovation while managing risks
- spaces to experiment and learn, with leadership backing and peer support
- transparency about when and how AI is being used, especially for external-facing communications.
“We need permission to experiment, but with clear boundaries.”
Focus group participant
Summary of key risks and barriers
Challenge | Impact | What’s needed |
| Informal use | Inconsistent practice, unmanaged risks. | Clear governance, policies, approved tools and sign-off routes. |
| Uneven access to AI tools | Capability gaps between organisations and teams. | Shared licensing models, organisational investment. |
| Skills and confidence gaps | Missed opportunities, poor prompting, unsafe use. | Role-specific training, prompt libraries, peer learning spaces. |
| Trust and ethical concerns | Potential for reputational damage, bias, misinformation. | Quality assurance, transparency and human oversight. |
| Cultural fear or resistance | Slower adoption, missed benefits. | Leadership engagement, positive framing of AI as support tool. |
The evidence suggests that NHS communicators are ready to embrace the benefits of AI, but they need the guardrails, guidance and support to do so safely, ethically, and confidently.
These findings directly inform the recommendations set out in the next section.
Conclusion and recommendations
AI is already playing a growing role in NHS communications, both through formal pilots and informal, shadow use. Teams are turning to AI tools to meet increasing demands with limited resources. In many cases, this adoption is ahead of policy, driven by practical need rather than strategic planning.
The research and the engagement that led up to the publication of our engagement paper in December 2024, highlights an urgent need for a structured, enabling response: one that builds confidence, protects trust and supports responsible innovation. While enthusiasm for AI is high, capability varies widely. Communications professionals working in the NHS want to use AI safely, ethically and effectively - with clear guidance and organisational support.
A consistent theme across the research is that AI should enhance, not replace, the human qualities that define NHS communications. Trust, empathy, tone and evidence must remain at the core. New skills - such as prompt writing, content verification and ethical risk assessment - will become essential for communicators at every level.
In response, the NHS Confederation and the NHS Communications AI Taskforce have identified five strategic priorities for action, each with a clear delivery route. These were set out in the engagement paper from December 2024, and the survey and focus groups we have undertaken on the back of the paper have reinforced the need for action across each of these five areas:
1. Develop a national AI operating framework for NHS communications
Informal use of AI is widespread. Without agreed guardrails, practice remains fragmented, with unmanaged risks to quality, trust and transparency.
Recommended actions:
- Set clear boundaries on acceptable use, including data input, human oversight, content review and publication.
- Provide clarity on when and how AI can be used safely (eg. for internal content, public messaging, or patient-facing materials).
- Address informal use by creating safe, sanctioned conditions for experimentation.
Planned delivery:
- This work will be addressed through the creation of a national operating framework for the use of AI in NHS communications. While not constituting formal NHS policy, the framework will provide critical guidance to local NHS communications teams developing their own official AI policies.
- Delivery will be coordinated by the NHS Communications AI Taskforce, with progress shared with the NHS Communications AI Network.
2. Develop an ethics framework
As AI becomes embedded in communications workflows, a strong ethical foundation is essential to retain public trust and safeguard NHS values.
Recommended actions:
- Establish ethical principles tailored to NHS communications, including transparency, accountability, human dignity and equity.
- Involve diverse voices, including frontline communicators and patients, in shaping the framework.
- Provide tools and scenarios to help teams apply ethical considerations in practice.
Planned delivery:
- Development of an ethics framework to run alongside the operating framework. A working group is being convened through the NHS Communications AI Taskforce, with 79 practitioners signed up to contribute at the time of publishing.
- This work will be progressed in summer 2025, with the aim of publication alongside the operating framework.
3. Promote the responsible and innovative of AI in NHS communications
AI can improve clarity, reduce jargon and enhance health literacy, but only when used thoughtfully and with audience needs in mind.
Recommended actions:
- Showcase real-world use cases that demonstrate impact and build confidence.
- Encourage use of AI tools that support inclusivity, plain language and accessibility.
- Position AI as a tool for enhancing, not replacing, the skills and values of NHS communicators.
Planned delivery:
- This work will be addressed through the creation of an online innovation and training hub. A working group has been convened to develop this resource, with 107 professionals expressing interest at the time of publishing.
- Progress will be shared via the NHS Communications AI Network, with development taking place over summer 2025.
4. Create a safe space for experimentation and shared learning
Teams want to test tools and workflows, but fear of failure or reputational risk can stifle innovation.
Recommended actions:
- Enable low-risk environments to trial tools, prompts and workflows.
- Promote shared learning across organisations, including both successful examples and lessons learned.
- Develop tools to support safe experimentation, including quality prompts, ethical checklists and peer feedback frameworks.
Planned delivery:
- Impact will be monitored via a monitoring and evaluation system, which will be developed alongside the innovation and training hub during summer/autumn 2025.
- A working group of 76 communications professionals has been convened to support this work.
- More generally, we have established an NHS Communications AI Network, which will connect peers to develop their use of AI by sharing practical insights and feedback on approaches that will help NHS organisations use AI to improve how they communicate with staff, patients and communities. This is being launched on 26 June. Around 200 communications professionals have already signed up to be part of the network. If you are a communications professional working in the NHS and would like to join, please sign up.
5. Provide role-specific training and guidance
Confidence and capability are uneven. Without support, AI’s benefits may be limited to a small number of early adopters.
Recommended actions:
- For operational teams: offer practical, task-based training in prompting, editing, tone-setting and verification.
- For senior leaders: deliver briefings on ethics, governance, risk management and leadership roles in AI adoption.
- Use a variety of formats, including:
- live demos
- NHS-specific case studies
- interactive prompt libraries
- downloadable guides and templates.
Planned delivery:
- The NHS Confederation will develop and coordinate this training programme, with launch planned for summer/autumn 2025, building on the findings of this survey and the forthcoming operating framework.
Summary of strategic priorities
Priority area | Delivery mechanism | Lead delivery organisation |
| Development of an AI operating framework for NHS communications (covering governance related issues) | National operating framework for the use of AI in NHS communications. | Taskforce and NHS Confederation, with regular updates provided through the in NHS Communications AI Network. |
| Development of an AI ethics framework for NHS communications | This will be a subset of the national operating framework for the use of AI in NHS communications. | Taskforce and NHS Confederation, with regular updates provided through the in NHS Communications AI Network. |
| Promote the responsible and innovative of AI in NHS communications | NHS AI Communication Innovation and Training Hub. | Specific working group, with oversight provided by the Taskforce and NHS Confederation. Regular updates provided via the NHS Communications AI Network. |
| Create a safe space for experimentation and shared learning | NHS Communications AI Monitoring and Evaluation System. | Specific working group, with oversight provided by the Taskforce and NHS Confederation. Regular updates provided via the NHS Communications AI Network. |
| Provide role-specific training and guidance | NHS Confederation. | National training offer and capability-building resources. |
This strategic programme aims to unlock the benefits of AI in NHS communications while protecting what matters most: trust, clarity, human connection and public confidence. The next phase of work will focus on delivering these actions and supporting communications teams across the NHS to innovate responsibly, build skills and share learning.
Appendices
HealthCommsPlus is a new creative communications agency brought to you by the NHS Confederation, providing a creative, professional and cost-effective service for NHS and public sector organisations across the UK.
Find out more about what we offer: www.nhsconfed.org/healthcommsplus
