We wanted to understand how AI is changing grantmakers’ practice, so this year we developed a new section of the survey to gather data on this. In total, 123 grantmakers responded to these new questions. This included 38 grantmaking trusts and foundations, as well as 85 organisations that make grants as part of their work. These include: Barts Charity, Northern Powergrid Foundation, Esmée Fairbairn Foundation, Wales Council for Voluntary Action (WCVA), Sussex Community Foundation, Waterside Community Fund, Cornwall Community Foundation, Sight Research UK, St Monica Trust, Devon Community Foundation and John Ellerman Foundation, as well as others that chose to respond anonymously.
AI adoption amongst grantmaking trusts and foundations
- Overall, the grantmaking trusts and foundations in our survey are further ahead with digital, data and AI.
- 71% say they are excellent, good or fair at using AI tools in everyday work (compared to 54% overall) and 83% are using AI tools in administration and project management (compared to 63% overall).
- 58% say they are at the stage of actively or strategically using AI in their organisation (higher than 38% overall).
- 80% believe that AI will make their work more efficient and impactful.
- 71% say they are excellent, good or fair at using AI tools in everyday work (compared to 54% overall) and 83% are using AI tools in administration and project management (compared to 63% overall).
- Like large charities, grantmaking trusts and foundations are most worried about the quality and accuracy of AI tools (61%), their energy use and environmental impact (55%), and the ethics or human rights concerns around using AI (48%).
How AI is changing the grantmaking process
Over half (58%) of grantmaking trusts and foundations have seen AI impact their applications. The biggest differences noted are an increase in volume (37%) and more repetition in applications (34%).
In this question we wanted to understand how AI is changing the grant application process from the perspective of grantmakers. Our discovery research with a small group of funders and stakeholders found that for some funders, AI use had led to an increase in the volume of applications, as well as a decline in the quality of applications. We wanted to find out if this was widespread. Our findings show the reality of the situation is more nuanced. At present, most grantmakers are uncertain about what the impact of AI has been on their processes.
- Overall, just over a third of grantmakers (35%), including trusts and foundations as well as charities making grants as part of their work, noted at least one way in which AI is having an impact on their funding applications.
- 18% have seen more repetition and similarity in applications.
- 17% have seen an increase in volume.
- 15% think that some applications are entirely AI generated.
- 10% are seeing more low-quality applications.
- 18% have seen more repetition and similarity in applications.
- Perhaps surprisingly, 62% of all the grantmakers did not see any changes with their applications, either because they had not seen a difference (25%), had not seen any of the differences we listed (7%), or did not know or feel comfortable saying (29%).
- Whilst our sample of grantmaking trusts and foundations is small (38), over half (58%) did note changes to their applications. The biggest differences noted are 37% saying they have seen an increase in volume and 34% seeing more repetition in applications. A small number of grantmakers shared observations about the impact AI is having on their applications:
“The response to certain questions is more often than not the same – for example, the expected outcomes.”
“We’ve only seen a few examples of AI-assisted or generated applications – not enough to make any sweeping statements concerning impact.”
“We are receiving more AI-generated applications that are low quality.”
- We can also see that grantmaking trusts and foundations are preparing for change associated with AI. 71% have or are developing an AI policy, whilst 48% are upskilling staff.
- These numbers indicate that AI is having an impact on grantmaking applications, but this is not as common as we thought and many grantmakers simply don’t know if there is an impact or have yet to see it.
What’s also interesting is that from a funder perspective, demand for funding of AI tools looks relatively low. Just 8% of grantmakers say that they are getting more applications to develop AI use and services, and only 2% are receiving more applications that include costs for AI tools. However, charities’ appetite to get AI use and tools funded is much higher, with 34% saying this is their biggest need, rising to 45% of large charities. It may be that there is a lag between how charities are using AI on the ground and what they then seek funding for, but nevertheless there is a disconnect here.
Do funders ask applicants to declare AI use?
Only 13% of funders ask applicants to declare or informally discuss their use of AI.
As some funders have provided guidance to grant applicants we wanted to understand how many had asked charities to declare use of AI in their applications.
- Two-thirds of funders (66%) have not done this.
- 21% do not know if this has been done.
- A mere 7% do ask charities to do this, with 6% asking informally.
Qualitative responses also indicate grantmakers are at early stages of considering this issue and monitoring the impact of AI on the grant application process:
“It has not been considered yet. The result is that we are guessing and bias creeps into assessments.”
“We haven’t asked yet, but maybe we should start asking.”
“We have only recently started collecting applicant data on AI usage so it’s too early to say.”
“It is clear when a bid writer is used, but not AI. We want them to be considered the same way, so we ask if AI has been used, and then also ask if a human has verified the information before submitting.”
How many charities use AI in funding applications, according to funders’ estimates?
More than half of funders (55%) do not know how many charities are using AI in their applications.
As we saw above, very few funders ask applicants whether they use AI to help write funding bids. Following this result, more than half of funders (55%) do not know and did not feel confident estimating what proportion of their applicants use AI.
Relatively small numbers estimate that the following proportions of charities are using AI in their applications:
- 19% of funders estimate that AI has not been used in their applications.
- 19% of funders estimate that less than half of charities are using AI.
- Only 7% of funders estimate that more than half of their applications use AI.
The answers to this question demonstrate that some funders are out of step with the reality of AI usage in the grant application process. Elsewhere in our report we highlighted how 45% of charities are using AI at the organisational level for grant fundraising. We can also see that this increases to 62% of global majority-led charities and 56% of neurodivergent-led charities, indicating that AI tools are making bid writing more accessible to these groups. If charities and funders had a shared understanding of this, useful learning could be exchanged on both sides.
How are funders using AI to support their assessments or decision making?
More than a third (36%) of funders avoid using AI in assessment and decision making. However, some are using AI to reduce admin, generate insights and summarise learnings.
In this question we wanted to find out how AI was changing funders’ assessment and decision-making processes. Funders appear to be at very early stages of AI adoption.
- 36% avoid using AI.
- 21% either don’t know or prefer not to say.
- A quarter (27%) of funders are using AI within their application processes. This includes:
- Reducing administrative tasks, such as summarising applications (15%).
- Generating insights from monitoring reports (12%).
- Summarising learnings from key calls and meetings (11%).
- Reducing administrative tasks, such as summarising applications (15%).
It is understandable that funders are being cautious about introducing AI into decision making and assessment. Less than 10% of funders are using AI for the following:
- Researching organisations and issues (8%).
- Identifying if applications are eligible (8%).
- Shortlisting and scoring against criteria (8%).
- Data analysis to identify trends and gaps in applications (7%).
However, some of the qualitative responses showed that funders are slowly starting to look into this.
“Not currently, but leveraging AI for due diligence checks and to help with drawing longlists from our growing database are both of interest to our grants team.”
“Also currently experimenting with using AI to help shortlist.”
Funders are right to be cautious in using AI for decision making. However, we do encourage funders to experiment with using AI tools, upskill themselves and learn from charities using AI. In the coming years, this will be essential to keep pace with charities’ use of AI and their funding applications, as well as the funding needs in the sector.
What are funders’ biggest hopes and fears for AI in terms of their funding and the issues they support?
In this open question, 47 grantmakers shared their hopes and fears for AI in terms of their grantmaking. Overall, 43% of responses were negative, 21% were positive, and 36% were neutral. Responses shed light on how funders are thinking about AI this year. Their responses cover the following themes:
Hope: More time for the human, relational side of grantmaking
The most common hope was that AI could take care of administrative work so that staff could focus on building relationships with applicants and communities, to understand them better. For some grantmakers, this was also linked to hopes that AI could open up funding access to applicants who have historically been disadvantaged by the application process, particularly those whose first language isn’t English.
“The biggest hope is that we can reduce the admin burden on members of the team so we can spend more time speaking with the applicant and understanding their needs.”
“My hope is that the use of AI can help to free up more time to devote to our relational funding approach and to work more closely with our local communities.”
“I hope the barrier to applications will be lowered, especially for those where English [is] not their first language, or those with great ideas too busy to otherwise submit.”
Fear: Declining quality and loss of authentic applicant voice
Some funders are worried that AI use is degrading the quality of grant applications, making them generic, lacking nuance, and stripping out the unique voice that allows assessors to understand what a charity really does. A related concern was that when many applications are edited by AI and read similarly well, it becomes harder to decide which projects deserve funding.
“More AI-generated applications, loss of charity ‘voice’. Lazy applications.”
“Biggest fear is that all applications will be AI generated and therefore all very high quality and it becomes difficult to differentiate which project deserves funding the most.”
“One of the biggest concerns is that the quality of application reduces further. Already we are seeing applications from organisations that might be good fits for us, but that are so poorly written that it is impossible to make a judgement.”
Fear: Exclusion of smaller, grassroots and less digitally mature organisations
While AI was seen by some as levelling the playing field, others worried it could do the opposite, excluding small or grassroots organisations without the digital capacity to engage with AI tools, or by widening the gap between AI-enabled and non-AI-enabled charities. This tension – AI as both equaliser and excluder – ran through some of the responses.
“Over-reliance on AI, both in grant applications and from funders…this could exclude extremely small organisations/changemakers who don’t have a digital footprint.”
“Widening inequality between organisations with access to AI and those without.”
“It is a shame that some groups who may benefit from our support are not able to put across their best nuanced application due to AI usage not truly reflecting their unique needs.”
Fear: AI in the review and assessment process itself
A distinct fear emerged around the use of AI by grantmakers, including whether reviewers might be using AI to assess applications (we know from our report data that low numbers of funders are doing this), the risks of bias being introduced into funding decisions, and the broader question of how to maintain integrity in the assessment process when AI is embedded on both sides.
“Unsure if our reviewers are using AI for their reviews. Will develop a policy and guidelines to make this clearer.”
“Bias when AI analyses applications to shortlist. Providing fake data.”
“There is a concern around the bias in AI. Both from how this might make its way into our work, but also whether there is a chance that we are feeding back into a biased system.”
Some grantmakers, including those funding environmental or social justice causes, raised concerns about the broader ethical and environmental implications of AI use. For some, this created a tension between embracing AI for efficiency and staying true to their values. One particularly thoughtful response also questioned whether AI is the right tool for the problems grantmakers face, arguing that the desire to use AI in monitoring or review is often a response to under-resourcing rather than a genuine technological need.
“Oh God there are so many fears. Environmental, ethical, war, and the erasure of critical thinking.”
“Whether open AI and generative AI can be used ethically (given the current politics and ownership of it) and how to be informed and make the correct values-driven decisions around its use in terms of ethics and environmental impact.”
“A concern is that the proposed uses of AI are sometimes a misguided response to challenges that are more rooted in longstanding resourcing issues than in technological issues. For example, a proposed use for AI is in reviewing monitoring reports but I think that this proposal is answering the wrong question. The question shouldn’t be ‘how do we use AI to make monitoring easier’ but ‘how have we gotten to the point where we don’t have the capacity to engage in a meaningful way with the information that we request from people, and if so, do we need to change our expectations around what we are requesting of others.”
These responses show that funders are alive to the possibilities and the risks of AI. These include the hope that AI could create more time for the human, relational work that grantmakers value most. Fears span quality, exclusion, assessment, ethics and environment. This tension, and our data about AI and grantmaking in this report, shows that whilst funders are at early stages with AI development, they are also trying to think through opportunities and challenges, much like the rest of the sector.