Progress with AI adoption
A recap on AI skills
Earlier in our report (see the section on ‘digital skills and capabilities’), we discovered:
- Skills to use AI tools in everyday work are low, but improving: Almost a quarter (23%) say they are poor at this, while 21% say they don’t do this. This means that 44% of charities either struggle with or are not using these tools. This is a significant improvement on last year, when almost two-thirds of charities (64%) were in these categories.
- Charities are struggling to keep up to date with emerging tech trends: Reinforcing this trend, 44% of charities are not keeping up with emerging tech trends. Just under a third (32%) of charities are poor at this and 12% don’t do this.
However, despite skills gaps, our extended questions on AI have uncovered increasing adoption and use of AI tools, particularly among large charities.
What stage are charities at with AI adoption?
Almost 8 out of 10 charities (79%) are using AI tools in some way, with a growth in charities actively and strategically using AI (38%, up from 25% in 2025).
AI usage has grown across the sector this year. The following proportions of charities are at these stages of AI adoption::
- Not at all: We don’t use any AI tools intentionally (20%, similar to 21% in 2025).
- Exploring: We are learning about AI tools and considering possible uses (41%, which is lower than 51% in 2025, because a higher proportion are actively using AI).
- Active: We use AI tools in some everyday tasks or areas of work (34%, up from 23% in 2025).
- Strategic: AI is widely used and part of our strategy or future plans (4% vs 2% in 2025).
Those in the exploring, active and strategic stages amount to 79% of charities, similar to the 76% last year. Large charities are further ahead here, with 92% in the exploring, active and strategic stages, compared to 74% of small charities.
Overall, it is clear that AI usage has increased, with 38% of charities now actively or strategically using AI tools in their everyday work, up from 25% in 2025.
Note: Our interim findings based on the first 240 responses showed AI usage at 88%. In the final weeks of the survey, we had an influx of responses from small charities and from those at an early stage with digital skills. The final result of 79%, based on the full survey sample, provides a more representative picture of AI adoption across the sector.
How are charities using AI day to day?
We’ve seen significant growth since 2025, with three in five charities (60%) now using AI for everyday tasks such as summarising meeting notes and drafting emails – up substantially from 48% last year. Only 12% of charities say they don’t use AI tools in their day-to-day work at all, down from 19% in 2025, indicating that AI tools have become embedded in a growing proportion of charities’ daily operations.
The top 5 ways charities are using AI day to day are:
- 60% are using AI for everyday tasks (e.g. summarising meeting notes, drafting emails), up significantly from 48% last year. This rises to 82% of large charities, vs 52% small.
- 57% are using AI to create documents, reports and presentations, up from 50% last year.
- 47% are using AI for research and information gathering, up significantly from 29% last year. This is the largest area of growth since 2025.
- 43% are using AI to generate ideas and creative thinking, up from 35% last year. This rises to 60% of large charities.
- 35% are using AI to shape funding applications, business proposals or pitches (a new option for 2026).
Large charities are using AI much more extensively, with 82% using AI in everyday tasks, 73% to create documents, and 60% to generate ideas and for creative thinking.
Charities’ use of AI for data analysis has grown significantly, suggesting this is now moving from an emerging to a more established use case. A quarter (25%) are using AI to analyse numerical data or spreadsheets, almost double the 13% that were doing this last year. This rises to 43% of large charities (compared to only 19% of small charities). Linked to this, later in our report (see the section on ‘data support needs’), we highlight how charities most want external support for using AI tools to analyse our data (47%).
This year we also asked about some emerging use cases, which were a little too nascent for us to justify gathering data on in 2025. For example, almost a third of charities (31%) are now using AI as a thinking partner to work through decisions. This indicates that charities are using AI in more nuanced ways to tap into additional capabilities, beyond productivity gains. A further 26% are using AI to review work for quality, bias, compliance or accessibility, suggesting charities are also starting to use these tools to strengthen their own outputs.
How are charities using AI organisationally?
2 in 5 charities (45%) are now using AI for grant fundraising, with administration and project management remaining the most common organisational use case at 63%.
The top 5 ways charities are using AI at the organisational level are:
- Administration and project management (63%, up significantly from 48% last year). This rises to 82% of large charities.
- Grant fundraising (45%, up from 36% last year).
- Social media (39%, similar to 36% last year).
- Communications and fundraising (36%, similar to 34% last year).
- Monitoring and evaluation (31%, up significantly from 17% last year). This rises to 42% of large charities.
The substantial growth in monitoring and evaluation stood out to us, suggesting charities are increasingly turning to AI to support impact reporting and analysis.
We also asked some new questions for 2026, and found that almost a quarter (23%) of charities are now using AI for knowledge management (e.g. for searching internal documents, creating organisational wikis), and 31% are using it for governance and compliance (e.g. policy drafting, board reporting). These findings suggest AI is starting to play a role in some of the more strategic and historically non-digital areas of charity work.
Interestingly for large charities, organisational uses of AI are focused on administration and project management (82%), with the second-biggest use case being monitoring and evaluation work (42%).
For small charities, AI is also mostly used for administration and project management (55%, lower than the 82% of large charities), closely followed by grant fundraising (52%, vs only 30% of large charities) and social media (43%, vs only 29% of large charities). This highlights how AI may be helping small charities to save time and add expertise in areas where they have less capacity (social media and bid writing).
However, charities are still only using AI in limited ways across some core functions. Only small proportions of charities are using AI for:
- Strategy, planning and predictive analysis (14%, up slightly from 9% last year).
- Technical digital work (e.g. coding, building tools) (13%).
- Learning and development (13%).
- Partnerships and business development (12%).
- Service delivery (9%, up slightly from 7% last year).
- Recruitment and HR (8%).
- Financial management (8%, up slightly from 4% last year).
The results across all of these areas are similar for both large and small charities. It’s notable that AI use in service delivery remains a very small proportion, despite the growth in AI use elsewhere. Charities are clearly exploring AI, but remain cautious in some areas that involve more risks.
Are charities providing AI advocacy and support services?
Less than 10% of charities are upskilling or advocating for their beneficiaries on AI.
This is a new question we asked to find out how many charities are looking at the impact of AI on those who are digitally excluded and people from marginalised communities, as these are concerns we hear from many charities. So far, only small proportions of charities are taking action in this area. We anticipate seeing this increase in the coming years.
- 9% are upskilling clients and organisations so they can use AI tools safely.
- Just 8% are influencing and advocating for safe and responsible AI development.
- Just 6% are researching AI use, experiences or impacts.
- 3% are offering support services for people harmed or at risk of harm linked to AI.
We recognise that some of these areas will not fall within many charities’ remits. However, given the number of charities that work with groups that are either from marginalised communities or which are digitally excluded, we would encourage more charities to think about how they can support beneficiaries through the societal shift that AI is creating. It is noteworthy that 29% of charities led by and for global majority communities say they are upskilling clients or organisations to understand AI and use tools safely, vs only 9% of charities overall. However, we appreciate that AI skills and digital inclusion in the context of AI are both new areas for charities, and that it will take a while for organisations to develop support offerings.
How are charities making use of AI agents?
28% of charities are either using, considering or testing agentic AI, rising to 39% of large charities.
In this new question we wanted to track how many charities are using AI agents, as these technologies have been much talked about since the last report. Last year just 11% were exploring AI innovations such as agentic AI automation tasks. This has grown to 28% this year, including:
- 16% currently considering or testing agentic AI (25% of large charities).
- 12% currently using agents (14% of large charities).
- Almost two-thirds of charities are not using agentic AI (64%, falling to 53% of large charities).
“We have built our own register sign-in app, automated registration forms added to a master spreadsheet linked to appsheet.”
“AI agent to search HR policies and procedures to provide quick answers to staff without needing to involve core team time.”
“Data analysis, reporting triggers calls to action on the workflow of our operations.”
Views on AI adoption
How are charities feeling about AI?
Almost two-thirds (63%) are excited about the potential for AI to increase their efficiency and impact. They feel less certain about whether their culture can enable innovation and the capacity for the organisation to respond to disruption.
This is one of several questions designed to tease out how charities feel about AI and the changes it will bring to their organisation.
Charities have mixed feelings about responding to AI:
- It is positive that 63% strongly agree or agree that they are excited about the potential for AI to improve their efficiency and increase their impact (similar to last year). This rises to 77% of large charities.
- However, when it comes to whether charities’ organisational cultures encourage experimentation and new ways of working, 32% are neutral and 23% disagree or strongly disagree. Both large and small charities feel the same about this. Only 41% believe their charity is well set up to adapt to AI.
- Whilst just over a third of charities (35%) agree that their charities are responding to the opportunities and challenges surrounding AI, 31% are neutral and 22% disagree or strongly disagree. This rises to 50% of large charities.
This suggests that whilst charities are excited about the potential of AI, they are less certain about whether their organisations’ culture and responsiveness will enable the change they would like to see. These findings are similar to 2025.
We compared responses to this question for different roles completing the survey. Whilst each role represents a different charity, they are representative of our sample, with identical proportions of each role responding on behalf of both large and small charities. We found similar responses from those in CEO and leadership roles, as well as staff and manager roles, so we combined these for analysis. We found that:
- CEOs and leaders are much more positive about AI adoption than staff, managers, and trustees.
- Nearly three-quarters (74%) of CEOs and leaders strongly agree or agree that they are excited about the potential for AI to make their work more efficient and impactful, compared to 60% of staff and managers. In comparison, 47% of trustees and board members feel excited about AI opportunities.
- Only 27% of trustees and 44% of staff and managers believe their organisational culture encourages experimentation and new ways of working with AI tools. Half (51%) of CEOs believe their charity has a culture conducive to AI adoption.
- Half (51%) of CEOs and leaders believe their organisation is responding to AI opportunities and challenges, but only a quarter of trustees (25%) and 39% of staff and managers feel the same way.
Benefits of using AI
What benefits have charities seen from AI?
In this new qualitative question, we wanted to understand the benefits that charities were seeing from AI. We had 433 responses and key themes included:
- Saving time on administrative and routine tasks, often freeing up capacity for relational, higher-value and strategic work
By far, the most common benefit charities cited was time saved on tasks that previously consumed significant staff capacity. This echoes our finding that 63% of charities are using AI organisationally for administration and project management, rising to 82% of large charities. Meeting notes, transcription, email drafting and routine administration came up repeatedly. Many respondents framed this as freeing up capacity for higher-value work. This included time with beneficiaries, frontline delivery and strategic thinking.
“AI drafting means first versions of grant applications, policies, reports and newsletters take minutes rather than hours, so staff can focus more on relationships, delivery and safeguarding.”
“We have seen a lot of staff time claimed back from e.g. typing up meeting notes, or call summaries – now these are done using Copilot.”
“Service delivery: exploring ways to improve service delivery by expediting staff-intensive manual tasks so the team can spend more time with service users.”
- Improving the quality and accessibility of writing
Many charities described AI as improving the quality of their written communications, particularly by making language clearer, more accessible and better tailored to the audience. This links to our finding that 57% are using AI to create documents, reports and presentations. However, several respondents also raised concerns about losing their tone of voice and producing generic written content.
“AI tools help speed up work, save time, improve grammar, and punctuation.”
“I use it regularly to turn my sometimes verbose language into plain English.”
- Supporting fundraising and bid writing
We know that 45% of charities are using AI to support grant applications and bid writing, particularly small charities (52%). These open responses highlight how AI is particularly beneficial for this, because of tasks like meeting strict word counts, structuring proposals, and getting started on the blank page. However, several respondents also flagged limitations, noting that AI-generated bids often need significant rewriting for quality, tone and accuracy.
“To edit grant application answers so that they fit the word limit but retain the meaning and the humanity of the original.”
“It can speed up the production of some work, e.g. bids and reports, however it rarely hits the right tone and organisational message first time and frequently needs significant tweaking before use.”
“Saving our unpaid time which grant funders do not respect.”
- Acting as a thinking partner and creative input
A number of respondents described using AI not just for productivity, but as a sounding board, critical friend or ideation tool. This echoes our quantitative finding that 31% of charities are using AI as a thinking partner to work through decisions.
“I also use it as a ‘critical friend’, to sense-check ideas and concepts, and to improve readability. AI helps me get started on projects when I have a mental block.”
“We are seeing people saving time through using AI for tasks across personal and team productivity, and a lot of staff are using it for critical thinking and ideation.”
“I have found AI tools transformative in terms of what I can produce and the quality of thinking I can bring to strategic work.”
- Supporting neurodivergent and disabled staff
A particularly powerful theme was the role of AI in supporting staff with disabilities or who are neurodivergent. As we can see in our profile on ‘charities led by and for neurodivergent people’, most of which are small, 65% are using AI tools in everyday tasks (compared to only 52% of small charities). In addition, 41% are using AI tools to review work for quality, bias, compliance or accessibility (vs only 26% in our wider sample). Several respondents described AI tools as transformative for their ability to work effectively, and this was true for senior leaders as well as more junior staff.
“I find the built-in note-taking and meeting transcription help me as an autistic person to process information, which is easier to do when it is written rather than spoken.”
“As a founder who is dyslexic, speech-to-text and AI writing tools have been particularly valuable in helping me work more efficiently and communicate ideas clearly.”
“As a neurodivergent Artistic Director, using structured tools significantly reduces the time and cognitive load involved in writing reports, defining complex tasks, and planning large-scale projects.”
- Doing things charities couldn’t do before
We know that charities are struggling financially, with 63% saying squeezed organisational finances are a significant challenge, while a lack of headspace affects half of charities (51%). A vital benefit of AI for some charities is that AI is helping them unlock new capabilities, particularly small charities and freelancers working with limited resources. This included data analysis, coding, building integrations and producing professional-looking content that previously would have required external expertise.
“Personally, AI is a game-changer. I am the digital lead for this charity, I am a freelancer. I couldn’t do the work I do and provide the level of service and depth of support to them without utilising AI.”
“It’s revolutionised our online offering. It’s saved us a huge amount of money on development of a new service platform for our beneficiaries.”
“Saving time and money, improving confidence to apply for funding and creating content I wouldn’t have felt confident doing before.”
Challenges and risks
What challenges, risks and practical issues are charities experiencing with AI?
We asked charities to tell us in their own words about the challenges, risks and practical issues they have experienced or anticipate when using AI tools, including reasons for deciding against AI use. 332 charities responded. The challenges they described fell into eight main themes.
- Concerns about accuracy, hallucinations and the need to verify outputs
By far the most common concern raised was the accuracy of AI-generated content. Many charities described AI tools as making mistakes, “hallucinating” information, or producing outputs that needed careful checking before they could be used. A common refrain was that the time saved by AI is partly offset by the time spent verifying its outputs.
“Hallucination. It is time-consuming to check everything generated by AI. It almost defeats the purpose of using AI.”
“AI has generated totally false and misleading data within research for reports that without prior knowledge and user diligence could have been reputationally damaging.”
“Have to keep reminding myself it’s just a machine and does not understand nuance and could present inaccuracies.”
- Data privacy, GDPR and safeguarding concerns
Concerns about data protection were a major theme, particularly for charities working with sensitive beneficiary information. Many respondents described feeling unsure about what data could safely be shared with AI tools, and some had decided not to use AI in client-facing work as a result. This was a particular concern for charities working with vulnerable people and children.
“I worry about sharing our service user case histories with AI as I don’t really want these shared on the wider web.”
“We work with vulnerable people and children – worried about the data protection side of it.”
“Nothing for myself but I don’t encourage others to use it as I am unsure they understand what data they can or can’t upload following GDPR.”
- Lack of training, skills and clear use cases
A recurring theme was the gap between AI’s potential and charities’ ability to use it well. Many respondents described not knowing where to start, lacking the skills to prompt AI effectively, or finding that without proper training, the outputs produced were poor quality. This often left charities reliant on a small number of digitally confident colleagues, with the rest of the team disengaged.
“We don’t know how AI could help us, and don’t know how to start.”
“Our volunteers are not very good with technology and it is all left to one person who does not have the time to develop everything.”
“Main challenges include limited staff skills and confidence, lack of formal training…Time and capacity to properly explore and implement AI safely is also a constraint.”
- Loss of authentic voice and the rise of “AI slop”
A particularly strong and emotive theme was concern about losing the authentic, human voice of charities in AI-generated content. Many respondents described AI outputs as generic, bland or obviously machine written, and worried about what this meant for their communications, their brand and the wider sector.
“I am concerned about the ethical and environmental impact but I also don’t want to lose our human heart and the skills that we bring to our work. I see a lot of organisations overly relying on AI, producing hard-to-read, nonsense slop and I don’t want that for us.”
“A freelance fundraiser using solely AI to write funding bids. Lost the charity’s voice and was very clearly written by AI.”
“Where it is used, it can sometimes subsume the individual’s own voice, or be used sloppily, leading at best to frustration, and at worst to substandard work.”
- Environmental and ethical concerns
A number of charities raised concerns about the environmental impact of AI, particularly its energy and water use. For some, this was framed as a tension between efficiency gains and the values of the charity, especially for organisations working on climate, sustainability or social justice issues. Ethical concerns extended beyond the environment to include the use of scraped creative content and concerns about AI replacing human work.
“AI is extremely destructive to the planet and we cannot afford the level of damage or water that it wastes while people need those resources.”
“As an arts organisation, we don’t use AI tools for generating art, creative writing, photo imagery or music. We consider this unethical because the models have scrapped existing artworks to generate these and we support the livelihoods of a team of freelance artists whose rights this has infringed upon.”
“There is a strong ethical concern about the environmental impact of AI and the climate crisis.”
- Cost, free versions and the difficulty of scaling AI use
Several charities described the cost of paid AI tools as a significant barrier, particularly when trying to roll out AI use across larger teams. Free versions were felt to be inadequate for serious use, but paid versions were often unaffordable, creating a catch-22 that limited what charities could do with AI.
“The cost of AI tools means that it is prohibitive when looking to roll out usage across the organisation (100+ staff).”
“Limited usage due to unpaid versions, sometimes produces random crap…Sometimes documents won’t download and then we have to waste another chat in our minimal allowance to ask again in another format.”
- Lack of policy and governance
A clear theme was the absence of organisational policies to guide AI use. Several charities described a situation where some staff were using AI informally while others held back, with no organisational direction on what was or wasn’t acceptable. Some charities were waiting for policy to be in place before using AI at all, while others described inconsistent practice as a growing source of cultural friction.
“Some of our staff use it and others are hesitant – we have no ruling currently to say whether AI can or cannot be used.”
“We are not using AI tools yet as we are waiting for AI policy and guidelines.”
“Early adopters can just paste AI output into emails leading to information overload – important to filter for efficiency rather than just present everything!”
- Charities deciding against AI use
A small but distinct group of charities described actively choosing not to use AI tools for ethical, mission or values reasons. This included arts organisations protecting creative livelihoods, charities prioritising human-centred work with beneficiaries, and individuals with moral objections to AI’s broader societal impact.
“I am morally and ethically opposed to the use and promotion of AI in automating daily tasks which can otherwise be performed by humans, and actively intend to avoid using them.”
“When applying for grant funding, we think it is preferable for our application content to be ‘human’ and from the heart.”
“As a literary arts organisation, we prefer to use, and that our beneficiaries use, human innovation in our interactions with each other.”
In summary, these responses show charities engaging with AI thoughtfully and with a clear view of the risks. Concerns are pragmatic (accuracy, time spent checking, cost), structural (training, policy, governance) and values-based (authentic voice, environmental impact, ethical sourcing). The variety of responses demonstrates the different perspectives about AI that exist across the sector. Charities need safe spaces to discuss these issues.
Moving forward with AI
What barriers do charities face when moving forward with AI?
More than half of charities (56%) see skills and technical expertise as the most significant barrier to AI progress.
People, not technology, are the greatest challenge charities face when moving forward with AI. A lack of skills, training and trust in AI are now the greatest hurdles to AI progress faced by charities, with concerns about colleague resistance also rising sharply. Concerns about data privacy remain a significant barrier alongside these.
The biggest barriers to charities’ AI progress are:
- Limited skills and technical expertise (56%). This has grown from 43% last year and is now the single biggest barrier to AI adoption. This is identical for both large and small charities (indicating it is a sector-wide problem, rather than capacity or resources).
- Lack of training or support to upskill (45%). This has risen from 39% last year, reinforcing the increasing focus on skills we are seeing elsewhere in the report.
- Data privacy, GDPR and security concerns (43%), unchanged from last year.
- Quality concerns about accuracy in AI-generated answers (40%), up from 33% last year, indicating this is growing in importance to charities. This rises to 47% for charities that are actively or strategically using AI tools. Quality concerns are also higher in large charities, with 60% citing this as a barrier.
- Lack of trust in AI tools (35%), up significantly from 15% last year. This is one of the most notable changes in this year’s data, suggesting that despite the growth in AI usage, charities do not fully trust the tools. Interestingly, for charities that are actively using AI, 32% still cite trust as a barrier. Charities supporting specific marginalised groups have more concerns about trust, cited by 43% of those working with global majority communities and with LGBTQIA+ groups. Trust is also a bigger concern for 44% of large charities, compared to only 32% of small charities, and this does not look to be dependent on the level of AI adoption.
The next tier of barriers includes a notable rise in concerns about colleague resistance and a reframing of cost barriers:
- Colleagues are hesitant or resistant to AI (34%), up from 20% last year. This suggests cultural and people-related challenges are becoming more of an issue in the sector.
- Free tools have limited features, usage or privacy (31%). We added this barrier to our list this year, expecting free tool limitations to affect small charities. However, the opposite is true, with 40% of large charities, compared to 28% of small charities, saying this is a barrier.
- The potential for bias and discrimination (31%), similar to last year. This rises to 43% of large charities. It also rises to 47% of charities working with LGBTQIA+ groups and 43% of charities working with global majority communities as target groups.
- Concerns about energy use and environmental impact (31%, up from 26% last year).
Some charities are being held back by:
- Ethical and human rights concerns (27%), similar to last year.
- Unclear on how it will help us (24%, new for 2026).
- Not a priority for our CEO/leaders (23%, up from 13% last year). This speaks to the leadership skills gap in AI we have seen previously in our report.
- Need to understand how AI will impact our beneficiaries (16%, new for 2026).
- Difficulty demonstrating AI’s value or return on investment (11%, down from 17% last year).
- Worries about impact on jobs and job security (9%), similar to last year (7%).
For those charities that are not using AI at all, the biggest barriers are:
- Limited skills and technical expertise (53%).
- Not a priority for our CEO/leaders (45%).
- Lack of training or support to upskill ourselves (43%).
- Unclear on how it will help us (e.g. no clear use case) (43%).
Skills, training, trust and colleague resistance now dominate the top of the list, replacing the data privacy concerns that led last year. This shows that growing people’s trust and skills will be vital to the next stage of AI development across the sector.
It’s notable that very few charities (9%) cite worries about job security as a significant barrier, reinforcing the finding from elsewhere in this report that AI is not currently being used to cut jobs in the charity sector. However, the increasing concerns about lack of prioritisation of AI by CEOs and leaders (23%, up from 13%) suggest that leadership engagement with AI is becoming a pressing challenge. This resonates with other trends we are seeing in the report.
How are charities affected by disintermediation?
11% of charities are developing content to be visible across different channels, with 8% reviewing the purpose and focus of their website.
We asked this new question based on our discovery research, which highlighted that disintermediation is an area of concern for some charities and digital agencies supporting them . Disintermediation is when AI tools provide information, advice and support about your cause, and in doing so disrupt how people find information from your organisation and interact with your services. If this is an issue you are worried about, CharityComms offers advice. The Patient Information Forum is also working on how AI is affecting health information. We only asked this question if charities were using AI in their organisation.
We were surprised to see that the number of charities affected by these issues was lower than we thought. However, this issue is clearly more prominent for large charities. The results are similar, even for charities who are confident they are making the most of their website, or at using website and social media analytics data. There are also no differences between charities led by and for specific marginalised groups in this area.
Regarding disintermediation, charities told us that:
- 48% say that this is not applicable or they don’t know, suggesting that many charities are not aware of this issue.
- 26% stated that they are not affected.
- Only 7% say that traffic to their website had fallen as a result of AI search tools. However, this rises to 19% of large charities.
- 11% of charities and 20% of large charities are developing content to be visible across different channels and to AI search tools.
- Only 8% of charities, but 20% of large charities, are reviewing the purpose and focus of their website because of the advent of AI.
However, charities are starting to prioritise this area of work, as we can see in the ‘digital communications and fundraising priorities’ section of this report, with 45% of large charities (and 29% of all charities) looking to develop content to improve visibility in AI search in the year ahead. We will continue to track the impact of disintermediation in future reports.
How many charities have AI policies?
47% of charities either have or are developing an AI policy.
- In total 47% either have or are developing an AI policy, similar to last year. This rises to 79% of large charities.
- This includes 18% with a formal policy, and 29% are developing an AI policy or have agreed good practice.
- 51% of charities do not have a policy, which is a concern. This rises to 62% of small charities, compared to only 21% of large charities.
What steps are charities taking to move forward with AI and manage risks?
44% of charities are not taking any actions to progress with AI and manage risks.
Worryingly, more than 4 out of 10 charities (44%) have not undertaken any of the actions listed to move forward with AI. This was the most popular option, and is a concern given the pace of AI adoption seen elsewhere in this report. This is predominantly small charities, with 55% saying they have not taken any action, compared to 15% of large charities.
The top actions charities are taking are as follows. As with last year, it is notable that even the most widely selected options are only being undertaken by around 2 out of 10 charities. There are also big differences between large and small charities here.
- Developing an AI strategy (19%), similar to last year. This increases to 39% of large charities. It is worth noting that 44% see developing or updating their strategy as a priority for the year ahead, cited in our section on ‘priorities and barriers’.
- Reviewing data protection, security and safeguarding (19%). This rises to 31% of large charities. However, we were surprised to see this had fallen from 36% last year, given ongoing concerns about these issues across the sector.
- Agreeing organisationally approved AI tools (17%, new for 2026).
- Making someone responsible for leading on AI (16%, new for 2026). This rises to 27% of large charities, compared to only 12% of small charities.
- Updating our risk register (15%, again worryingly this is down from 23% in 2025).
Very few charities are taking some of the most important governance steps. Only 7% have started reviewing AI use and risks regularly at board meetings, only 6% have conducted risk assessments for AI use cases (e.g. bias, misinformation), and only 6% have started getting their data ready for responsible AI use. Just 4% are partnering with corporates or the public sector to share expertise, similar to 9% last year. The latter could be a missed opportunity, as other sectors could benefit from the expertise of charities that have valuable insights into their communities and causes.
At a time of rapid adoption we are concerned to see how many charities are not taking steps to either progress or manage risks. Furthermore, only 10% of charities are planning for future AI developments and their impacts (rising to 21% of large charities).
How are charities planning for the future of the workforce?
Almost 3 out of 10 charities (29%) are upskilling staff and volunteers on AI, which reinforces the focus we are seeing on skills in this year’s report.
In this new question we wanted to gather data on a topic that has been dominating headlines in other sectors: how AI is impacting the future of the workforce.
The default position is that almost half of charities (45%) are not taking any of the actions listed.
The top areas where charities are most likely to be taking action are:
- Upskilling staff and volunteers on AI (29%), which mirrors a focus on upskilling that we have seen in the rest of the report. This increases to 42% of large charities.
- Similarly, 1 in 5 (20%) are upskilling leaders and boards.
Small proportions of charities are taking action in the following ways, with only minor differences between large and small charities:
- 10% are planning for future impacts on staff and roles (16% of large charities).
- 8% are considering AI literacy in recruitment.
- 8% are reshaping staff roles and responsibilities.
It is also worth noting how few charities are taking more structural action. Only 6% are integrating AI skills into role descriptions or recruiting staff with AI expertise, just 5% are reallocating resources as AI saves time on tasks like administration, and only 2% are reallocating staff to different roles or teams. Reassuringly, and contrary to headlines in other sectors, only 1% of charities are reducing staff in response to AI. Taken together, these findings suggest that while charities are not using AI to cut jobs, they are also not yet redesigning what organisations and jobs could look like. Looking at the changing trends elsewhere in the report it may be that this is happening organically, driven by staff, rather than from the top down.
In summary, these numbers show that charities are very much at the start of the journey of planning for the future of the workforce. However, given the numbers of charities now using AI, we hope that more charities will offer their staff, leaders and boards opportunities to upskill, and plan strategically for inevitable workforce changes. This will ensure that charities are able to build, attract and retain the skills and talent they need.
What sector-wide initiatives would help charities move forward with AI adoption?
Upskilling staff, volunteers and leaders is the most important sector-wide effort charities would like to see.
The top 5 sector-wide initiatives both small and large charities need are:
- 60% want training for staff and volunteers.
- 57% want training for leaders and trustees, similar to last year.
- 45% want sector guidance on responsible AI use and examples, down significantly from 69% last year. However, large charities in particular would value this (58%).
- 44% want guidance on developing AI policies and governance.
- 41% are interested in AI tools developed by and for the sector, rising to 56% of large charities.
Large charities would also value more peer learning and coordination between charities to share use cases (48%), as well as insights about AI’s impact (48%).
Charities also want funders and regulators to take the following actions:
- Just over a third (36%) want more funding for AI tools and adoption.
- 29% want funder plus support or funded specialists to provide support. This rises to 45% of global majority-led charities, which say this is what they need most.
- 23% want more funder and regulator guidance on their policies.
