Why Copilot Alone Isn’t Enough – and What UK Fundraisers Are Missing

I’ve been having the same conversation again and again.

A fundraiser gets in touch, excited about what AI could unlock for their charity… then adds, slightly sheepish:

“But we’re only allowed to use Copilot.”

I know why it happens. IT teams have done their due diligence. Microsoft is already the default. Copilot feels safe, sanctioned and corporate-approved. From a risk perspective, that’s entirely reasonable.

But it’s also limiting.

Relying on Copilot alone is a bit like telling a chef they’re only allowed one knife. You’ll still get food on the plate – but you won’t get the best results, and you’ll work harder than you need to.

Copilot is genuinely good at what it does. The problem is what it wasn’t designed to do.

Here are five opportunities UK charities are missing when Copilot becomes the only AI tool fundraisers are allowed to use.

1. Fundraising Isn’t Mechanical – It’s Emotional

Copilot excels at productivity. Summaries. Drafts. Analysis. It’s brilliant at the mechanics of work.

But fundraising isn’t mechanical. At its heart, it’s about emotion, meaning and human connection.

In our AI-versus-human thank you letter research, the highest-scoring letters didn’t come from generic tools. They came from custom-trained models fed with gratitude psychology, stewardship research and examples of exceptional supporter care.

Tools like Anthropic’s Claude tend to produce warmer, more nuanced writing.. OpenAI’s ChatGPT allows deep customisation through tailored project GPTs. Google’s Gemini has Gems - custom projects that you can train on your content.

The missed opportunity: charities settle for serviceable supporter comms when emotionally intelligent content is within reach.

2. Copilot Has a Prompting Ceiling

Copilot is designed for general Microsoft Office users. It’s intentionally frictionless.

That sounds great – until you realise it limits how far expertise can take you.

Advanced prompting techniques, such as role-play, structured reasoning, few-shot examples, custom instructions, transform AI from a basic assistant into a genuine thinking partner. With ChatGPT or Claude, those techniques are rewarded. With Copilot, they’re largely flattened.

At Thread, we use specialist AI setups trained on our own fundraising library we’ve created over the last 20+ years. We use proven techniques, frameworks and examples to create custom solutions. Copilot simply isn’t built for that depth.

The missed opportunity: fundraisers who invest in AI skills can’t fully use them. To stretch the chef analogy, it’s like having a pro chef that you only allow to use a microwave to cook with.

3. Innovation Moves Faster Elsewhere

AI development is moving at extraordinary speed.

ChatGPT releases new models and features at pace. Claude introduced long-memory “Projects”, “Skills” and MCP connectors (making it easier for tools to ‘talk’ to each other.

Tools like Perplexity blend search with reasoning. NotebookLM turns research into reports, courses and podcasts..

Copilot is improving, but it’s constrained by enterprise rollout cycles and Microsoft’s priorities.

The missed opportunity: by the time features reach Copilot, charities restricted to it have already lost months of productivity and creative advantage.

4. Different AIs Are Good at Different Things

In real fundraising work, no one tool wins everything.

In one recent case study we used:

• ChatGPT for ideation and first drafts

• Claude for refining donor-facing copy

• Gemini for Excel-based data analysis

• Automation tools to stitch everything together

Each tool had a clear role.

Back to the chefs - using only Copilot is like running a professional kitchen with only limited equipment. Possible? Yes. Optimal? No.

The missed opportunity: fundraisers can’t match the tool to the task – so outputs suffer and time is wasted.

5. Copilot Doesn’t Build AI Literacy

There’s a bigger, longer-term issue here.

The fundraisers who will thrive over the next five years aren’t just using AI, they understand it. They know how prompting works, where models fail, and how to adapt their thinking.

Copilot’s seamless integration means you get answers without learning much about what’s happening under the bonnet.

Working directly with other tools builds transferable skills:

• How structure affects outputs

• When reasoning breaks down

• How to design prompts for consistency

The missed opportunity: fundraisers aren’t developing the AI fluency that will soon be core professional literacy.

Where Copilot Does Shine

None of this is anti-Copilot. Used well, it’s genuinely powerful.

Three areas stand out:

Excel data analysis: cohort analysis, pivots, cleaning messy data

Meeting and document summaries: especially in Teams and Word

Outlook email drafting: quick, contextual first drafts you then personalise

If Copilot is your only option, focus your energy there. You’ll get real value.

The Path Forward Isn’t Either/Or

The best answer isn’t choosing between Copilot and other tools. It’s building access to the right tool for the job, without overcommitting to a single model or vendor.

For many charities, that starts with sensible, graduated permissions:

• Copilot for all staff – day-to-day Microsoft integration, summaries and Excel analysis

• Specialist AI tools for content and strategy – with clear rules about data (no supporter data, no confidential financials)

• Centrally governed setups that give flexibility without increasing risk

This is where tools like TypingMind and OpenRouter come in.

Together, they offer a genuinely practical middle ground.

TypingMind provides a secure, user-friendly interface that feels familiar to non-technical teams. OpenRouter sits behind it, giving access to multiple AI models, Claude, GPT-4-class models and others, on a pay-as-you-use basis.

No long-term lock-in.

No need to “pick a winner” model too early.

No paying premium licence costs for tools you only use occasionally.

Fundraisers can:

• Use Claude when warmth and nuance matter

• Switch to GPT-style models for ideation or structured problem-solving

• Experiment with new models as they emerge

• Match cost to actual usage, rather than overcommitting budgets

For IT and leadership teams, this approach often reduces risk rather than increasing it. Fewer shadow tools. Clear governance. Central oversight. And the flexibility to adapt as the AI landscape changes, which it will.

Better fundraising and supporter experiences need the freedom to use the right tools, thoughtfully, ethically and in service of the people who make our work possible.

If this resonates, I’d love to hear your experience.

Are you Copilot-only? Where does it help – and where does it hold you back?

These conversations matter. The future of UK fundraising will be shaped by how thoughtfully we use these tools, together.

Join our free webinar on Wednesday 21st January at 1pm to show how we’re using these tools to create campaign ideas quickly and efficiently.

Want to explore how you can create your own AI-powered ‘fundraising brain’ to help your team? Get in touch for a chat.