Watch our Video
AI is already being used by many charities in writing grant applications but in careful, limited and very human ways. Findings from the Charity Excellence Grant Making Trust & Foundation Survey 2026 show a sector that is experimenting, learning, and setting clear boundaries around how AI should (and should not) be used in funding bids.
AI use is now mainstream, though not universal. Around 64% of charities have already tried using AI or use it regularly when writing grant applications, while a further 15% plan to use it but haven’t started yet. Only 14% say they do not intend to use AI at all. This shows that AI is quickly becoming part of everyday fundraising practice, particularly for time‑pressed teams.
Crucially, frequent use does not mean wholesale handover to machines. Most respondents describe supportive use, not automation of grant writing.
“I approach funding bids by writing the full response and then tidy up using AI, e.g. to meet word counts and to create better flow and reduce repetition.”
“It is helpful in trying to help understand what the question is asking as I have dyslexia.”
Charities who use AI report clear practical benefits:
Common uses include reducing word counts, improving flow, checking whether questions have actually been answered, organising ideas, and refining tone – rather than writing bids from scratch.
“The biggest impact it's had on my work is answering questions about linking our work to other bodies' plans and strategies. For one application, it probably saved me a day's work.”
“You absolutely cannot just dump AI generated responses into a funding bid.”
The risk that seems to be most often cited is of a rapid growth in very poor quality, generic grant application volumes, similar to that experienced in recruitment with job applications.
Views on ethics are broadly positive but cautious. Around 60% agree that using AI for bid writing is ethical, provided it is used responsibly. A similar proportion feel confident managing risks such as plagiarism, data protection, and generic content but this confidence is not universal.
Many respondents stress the importance of:
“Every single piece of output needs to be checked for accuracy. The limitations of the tool must be recognised as well as the potential.”
There is strong agreement that AI‑generated content must always be edited, humanised and sense‑checked.
“Lottery spoke about using AI at the Impact Summit. They said they are aware that it is being used and were absolutely ok with applications being supported by AI.”
Feedback about AI use from grant makers is mixed, inconsistent and sometimes contradictory. Some charities report explicit push‑back, particularly from smaller trusts and foundations, while others have heard more neutral or pragmatic messages.
“National Lottery told me they can tell an AI‑influenced bid a mile away – they don't seem to like it!”
However, other respondents report funders being open to AI as a supporting tool, provided authenticity is preserved.
“On the whole, the grant makers I have spoken to aren't generally keen on applicants using AI to write funding bids. This is especially true of many small trusts and foundations.”
Across responses, one message comes through clearly from funders: passion, individuality and real understanding of the community matter more than polished wording.
“Make sure that you tell the story with passion and enthusiasm, ask it to use your language and ensure AI doesn't make it bland.”
There is also concern that increased AI use may be affecting the wider funding environment.
“I worry that identical AI‑written applications make it easy for a funder to rule out huge swathes of bids.”
“AI is a tool not the job. You need to personalise to your charity and add emotion.”
Across the survey, a strong consensus emerges:
“I use AI to improve my content but quite often reject its suggestions and rewrite in my own voice.”
Pure AI‑written funding applications are often easy for experienced reviewers to spot because they tend to be generic and lack organisational voice. However, where AI is used as a drafting or editing aid and the content is then properly reviewed, rewritten and humanised, there is currently no reliable way, technical or human, to identify AI involvement. Evidence shows both detection tools and reviewers struggle to distinguish hybrid, human‑edited AI work from fully human writing.
Some grant makers appear to think they can identify AI drafted applications when what they are actually doing is spotting bad AI applications. We are also concerned (but have no evidence) that individuals who learned English outside the UK and who use US spelling (such as asylum seekers) may be unfairly penalised by grant makers who assume their application is AI drafted, when it is not.
The ACF report Foundations in Focus (2025) stated that in 23/24 grants increased by 12% and that applications surged dramatically, with some foundations reporting up to 400% more. This was attributed to inflation, cost-of-living pressures, and increased demand for charity services, which have surged massively in recent years.
However, we do not think this is the whole picture. The factor we think needs to be added in is Government funding. The Government has been cutting departmental budgets year on year as well, not least in mid 2023, when our £1bn Funding Cuts report predicted the collapse in fundraising resilience that became the longest crisis so far.
Unlike fundraising, Government funding does not increase again as the economy recovers. What we think we're seeing is charities increasingly trying to replace lost public sector funding with charitable funding, making an already bad situation worse.
Our survey has identified no suggestion of widespread use by charities to generate large volumes of low quality bids.
For most charities, the question is no longer whether to use AI, but how to use it well. The evidence suggests careful, transparent and limited use that is grounded in the charity’s own knowledge and language, without losing what makes each charity distinctive.
Everyone wants small and marginalised groups to have equal access but the need for good written English and an understanding of what grant makers want, often stacks the odds against them. Particularly, those with poor educational attainment, learning difficulties or for whom English is a second language. Equally, the number of bids that a charity needs to submit has drastically increased in recent years. Most of the evidence is anecdotal but widely accepted. This is creating more work for grant makers and a very significant burden on hard pressed front line charities.
AI offers us a huge opportunity to mitigate both issues but contradictory messaging by grant makers risks having a chilling effect on its use. What is required to make AI a useful tool for everyone is:
We have published best practice guidance for grant makers in using AI and also for grant applicants, and our Charity Excellence Learning, free online AI training includes a course on effective bid writing using AI.
Source: Grant Making Trust & Foundation Survey 2026. Quotes reproduced verbatim from respondents. Researched by AI – verified by a human. Added 9 Apr 2026.
A registered charity ourselves, the CEF works for any non profit, not just charities.
Plus, 60+ policies, 8 online health checks, the Quality Mark and the huge resource base. Our AI Ready programme and Charity Excellence Learning free online AI training courses, give non profits everything they need to make effective use of AI and stay safe.
Find Funding, Free Help & Resources - Everything Is Free.
Yes. Most charities now use AI when writing grant applications, and survey evidence suggests this is widely viewed as acceptable when done responsibly. AI is generally seen as an assistive tool, not a replacement for human judgement. Ethical concerns arise only when AI is used without oversight, accuracy checks or human input. Responsible use means the charity retains control of content, accuracy and tone, and ensures the application genuinely reflects its real work and values.
Charities mainly use AI to support and refine their own work rather than to generate bids from scratch. Common uses include clarifying questions, improving structure and flow, reducing word counts, checking alignment with funder criteria, and organising ideas. The core narrative is usually written by a person, with AI used afterwards to tidy, refine or sense‑check the draft. This supportive approach helps save time while preserving authenticity.
Survey respondents say not with strong agreement that using AI to write whole applications without human input leads to poor results. Fully AI‑written bids are often generic, lack organisational voice, and fail to reflect lived experience or local context. Most charities report deliberately avoiding this approach and instead using AI only after drafting their own content, ensuring control over substance, tone and accuracy.
Pure, unedited AI‑written applications are often easy for experienced reviewers to identify because they feel generic and lack specificity. However, when AI is used as a drafting or editing aid and the content is then properly rewritten and humanised, there is currently no reliable technical or human method to identify AI involvement. Evidence shows both detection tools and reviewers struggle to distinguish high‑quality hybrid work from fully human writing.
There is no evidence of widespread rejection solely because AI was used. Where applications are unsuccessful, this is usually due to poor quality, generic language or lack of authenticity. In many cases, what funders believe are “AI applications” are actually weak applications. Funders consistently emphasise that they care about clarity, insight and genuine understanding of community need, not whether a tool was involved in drafting.
The biggest risk is producing generic, bland applications that fail to stand out. Other risks include factual errors, misinterpreting funder criteria, data protection concerns and losing the charity’s authentic voice. These risks increase when AI output is used without review. All are manageable if charities treat AI as a support tool and ensure every output is checked, edited and rewritten by someone who understands the organisation and its work.
There is clear evidence that grant application volumes have surged in recent years, but this is mainly linked to inflation, cost‑of‑living pressures, rising service demand and cuts to public sector funding. While concerns exist about AI enabling mass‑produced bids, our survey found no indication that charities are using AI to generate large volumes of low‑quality applications. Most charities remain cautious and selective in their use.
Used well, AI can reduce barriers for small charities, people with dyslexia, learning difficulties, or limited fundraising experience, and those for whom English is a second language. However, there is concern that some reviewers may incorrectly assume non‑UK spelling or phrasing indicates AI use. While there is no firm evidence of widespread bias, this risk underlines the need for fair assessment and better understanding of how AI‑assisted writing actually works.
Funders consistently emphasise passion, individuality and real understanding of the community being served. They value clarity and honesty over polished language. Applications that reflect lived experience, local knowledge and genuine organisational voice perform better than generic submissions. AI can support clarity and structure, but funders expect the charity’s values, insight and judgement to remain visible throughout the application.
Good practice means drafting the main content in your own words, using AI to refine and improve it, and checking every output carefully. Charities should remove generic language, verify facts, and ensure the final application clearly reflects their real work and voice. AI should support decision‑making and writing quality, not replace judgement, experience or accountability.
For most charities, the question is no longer whether to use AI, but how to use it well. Careful, transparent and limited use can save time, improve clarity and widen access without undermining trust. Clear guidance from funders and better AI training for charities are essential to avoid mixed messages and ensure AI becomes a levelling tool rather than a new source of inequality.
