Digital isn’t neutral: How 2025’s charity digital trends are deepening inequity

10^th December 2025

Ellie Hale

Quick links

Many organisations are investing in tools before investing in justice for the communities they serve. Here are some actions organisations can take in solidarity with communities already leading this work.

The 2025 Charity Digital Skills Report surfaced a widening digital equity gap that barely registered in its AI-dominated headlines. At Catalyst we think that in a year defined by polarisation, exclusion and rapid technological shifts, it’s more important than ever to ask: what gets funded, who gets listened to, and whose needs shape our digital future?

The data tells a consistent story: charities led by minoritised groups are doing some of the most inclusive, values-aligned and ethically cautious digital work, while facing the biggest structural barriers to funding and power.

Here are five underreported tech justice issues from this year’s report, and why they matter for anyone making decisions about funding, design or governance.

1. Co-design is shallow and minoritised people are structurally excluded from decision-making

What the stats say

72% of charities say they are co-designing their services with users. But only 15% are researching with users from diverse communities to a great extent.
66% of charities say that their services are developed by diverse teams. But only 15% do this to a great extent.
Only 28% of charities see accessibility, diversity and inclusion as a top priority in digital services.

What it means

This is tick-box inclusion, not shared power. ‘Diverse teams’ mostly aren’t, and minoritised communities are still largely excluded from shaping and owning digital solutions. As a result, those digital services are more likely to keep reproducing inequity.

What to do next

For leaders and delivery teams: Set a minimum standard that all new or redesigned services include research and testing with people most affected, documented in discovery and evaluation.
For funders: Require evidence of deep community involvement (not just “consultation”) in funding applications and budget lines for participant time, access needs and translation.

2. Equity work is led by minoritised charities but starved of funding

What the stats say

Black-led charities are significantly more likely than the sector overall to say their digital services are informed by research with diverse communities to a great extent (31% vs 15%), reaching diverse and minoritised communities (34% vs 20%) and monitoring diversity and inclusion (26% vs 17%).
Accessibility, diversity and inclusion are a top digital priority for 53% of disabled/d/Deaf-led charities, 43% of neurodivergent-led charities and 40% of Black-led and LGBTQIA+-led charities, compared with 28% overall.
Finding a funder for digital projects is the biggest challenge for 27% of charities overall, but that rises to 44% of Black-led charities (nearly double the average), 34% of lived-experience-led charities, 33% of disabled/d/Deaf-led organisations and 35% of neurodivergent-led charities. The same disproportionate pattern appeared in last year’s report.

What it means

This is extractive progress (taking without replenishing) – the groups best positioned to deliver just, inclusive tech are the least resourced to do so, while shouldering the lion’s share of ethical labour.

What to do next

For funders: Ring-fence funding for Black-led, disabled/d/Deaf, neurodivergent and LGBTQIA+ organisations leading digital work, and simplify application processes that currently exclude them.
For leaders: Build long-term partnerships with minoritised-led organisations, treating them as paid strategic partners and co-owners of digital initiatives rather than occasional “advisors.”

3. Those closest to harm are doing the most to avoid harmful AI and tech risks.

What the stats say

65% of neurodivergent-led charities are worried about the implications of AI and 47% are actively avoiding AI where it could cause harm, compared with 36% overall.
49% of LGBTQIA+-led charities see the potential for bias and discrimination in AI as a barrier, and 49% cite ethical and human rights concerns (vs 32% and 29% overall).
Neurodivergent-led charities are more likely to say their services are developed by diverse teams (29% vs 15% overall) and inclusive of users (40% vs 33%), and they prioritise suppliers who involve people with lived experience (49% vs 34%).

What it means

These organisations are critical stewards of ethical tech – holding accountability despite having the fewest resources. This is a common liberatory pattern: those closest to harm are most committed to avoiding it.

What to do next

For leaders: Invite neurodivergent, LGBTQIA+ and other minoritised-led organisations into AI governance structures, advisory groups or steering committees with real decision-making power (while being mindful of the value exchange).
For delivery teams: Use their risk awareness as a template – document harms, biases and red lines for your own AI use, and build opt-outs or alternatives into services.

4. Digital capacity is weakest in the regions that need it most

What the stats say

Charities in Northern Ireland, the North of England and the Midlands report lower digital maturity, weaker funding and reduced AI uptake than many other regions.
Supporting clients with devices, data or digital skills is an organisational priority for 25% of charities in the North of England, 22% in the Midlands, 25% in Scotland and 22% in Wales, compared with 16% overall.

What it means

Digital progress is mapping onto geographic inequality. Tech capacity is weakest where need is greatest – again deepening structural injustice, this time across regions.

What to do next

For funders: Prioritise targeted investment, capacity-building and peer networks in Northern Ireland, the North of England, the Midlands and other under-resourced regions, not just major cities.
For leaders: Share resources and practice across regions by joining or forming cross-region learning groups, and budget for travel, translation and accessibility so smaller orgs can participate.

5. AI enthusiasm is outpacing strategy and inclusion

What the stats say

- 76% of charities are using or exploring AI tools, up from 61% in 2024.
- Only 44% have a digital strategy in place, down from 50% last year.

Only 28% of charities see accessibility, diversity and inclusion in digital services as a top priority.

What it means

Innovation is accelerating on shaky, inequitable ground. The disparity between AI enthusiasm and strong strategic foundations was picked up by several report commentators, but they neglected to include the crucial inclusion stat, which hides the extent of the problem. Without prioritising ethics, participatory governance and lived-experience leadership, the sector risks automating exclusion at scale.

What to do next

For leaders: Put in place or update a digital strategy, including commitments on accessibility, inclusion and governance, before scaling AI tools any further.
For delivery teams: Slow down AI rollouts that lack clear user benefits or bias safeguards, and prioritise improvements to core digital services and accessibility first.

The takeaway for anyone working in digital and data

We cannot allow a sector rooted in social justice to recreate systems of oppression through its digital choices. AI clearly offers huge opportunities for organisations, but it’s our values that should direct our approach and priorities. Tools should always follow and support, not replace, the harder work of deciding who holds power over digital choices.

If we want our digital work to be truly liberatory, it has to meaningfully centre those most affected by injustice – not just as service users, but as people with real power and voice in shaping what gets built.

What we can do in our organisations

Set a quarterly target for inclusive research: commit only to projects informed by research with diverse communities “to a great extent.” Accept that this may require more time, effort and budget and plan (and fundraise) accordingly.
Prioritise tech justice principles in strategy: make accessibility, diversity, inclusion and equitable participation a top digital priority at the same level as AI pilots. While ensuring it’s part of those pilots too.
Strengthen governance: ensure a current digital strategy exists before scaling AI. Conduct AI ethics reviews and harm assessments before adoption, put in reflection checkpoints and mitigation plans. Respect and recognise the perspectives and leadership of communities closest to AI-exacerbated harms.
Fund lived-experience leadership: ring-fence budget for minoritised-led partners and compensate co-design participants fairly. Catalyst’s reciprocity policy might be a helpful starting point.
Funders: address systemic and regional gaps: allocate additional support to minoritised-led charities and those based in Northern Ireland, the North of England and the Midlands. And given their greater focus on tech justice principles and practices that benefit everyone, organisations led by minoritised groups should be supported to lead and shape the sector’s digital future.

—–

Appendix: Key tech justice stats that deserve attention

72% of charities say they are co-designing their services with users either to a great extent or some extent. However, only 15% of charities say their services are informed by research with users from diverse communities to a great extent.
66% of charities say that their services are developed by diverse teams, which has increased from 45% in 2024. However, only 15% say they do this to a great extent.
Commitment to diversity and inclusion is very important to 42% of charities when choosing a supplier. This has more than doubled from 19% last year. This increases to 45% of small charities, compared to only 36% of large charities. It also increases to 62% of black led charities, 66% of disabled and d/Deaf led organisations and 68% of LGBTQI+ led charities. 34% say that how the supplier involves users and people with lived experience is very important, which has doubled from 17% in 2024. 55% of infrastructure organisations say suppliers’ commitment to diversity and inclusion is very important in their selection.
Finding a funder who supports a digital project is the biggest barrier for 27% of charities (28% of small charities and 25% of large charities) and has increased from 21% in 2024. Funding for digital projects is a bigger challenge for 44% of black led charities, as well as 34% of lived experience led charities, 33% of disabled and d/Deaf led organisations and 35% of neurodivergent led charities.
Black led charities are more likely to say their digital services are (to a great extent):
- Informed by research with diverse communities (31%, vs 15% overall)
- Reaching diverse and minoritised communities (34%, vs 20% overall)
- Monitoring diversity and inclusion (26%, vs 17% of the main sample).
35% of Black led charities are seeking funding for AI (vs 13% overall). Black led charities say that co-ordinated AI investment from funders would make the biggest difference to sector adoption (53%, vs 35% of the main sample). 44% want more insights about how their communities are using AI (vs 25% overall). 40% are developing an AI policy. A higher proportion of black led charities are avoiding AI in areas where it could cause harm (47%, vs 36% overall).
The challenge of finding a funder for a digital project is affecting black led charities (44%) disproportionately. This compares to 27% of charities overall and 28% of small charities. The same results were uncovered last year (41% of Black led charities struggle to find a funder who will support a digital project, compared to 21% of the main sample and 25% of small charities)
Neurodivergent led charities say their digital services are more likely to be developed by diverse teams (29%, vs 15% overall) and inclusive of users (40%, vs 33% overall). Their digital suppliers are chosen on the basis of their commitment to diversity and inclusion (61%, vs 42% overall), how they involve people with lived experience (49%, vs 34% overall) and diversity in their team (28%, vs 17% overall). 65% are worried about the implications of AI and 47% say they are avoiding AI in areas where it could cause harm (vs 36% overall).
65% of LGBTQIA+ charities stopped using or reduced their use of X (formerly Twitter) this year (vs 51% overall) and 11% reduced their use of Meta platforms. 42% have changed their social media strategy (vs 25% overall). 29% have changed how they post content and 18% have spent more time tackling misinformation (vs 5% overall).
Barriers to adopting AI tools are distinct for LGBTQIA+ led charities:
- Data privacy, GDPR and security concerns (53%, vs 43% overall)
- The potential for bias and discrimination (49%, vs 32% overall)
- Ethical and human rights concerns (49%, vs 29% overall).