How to assess whether an AI tool is right for your charity

There are some key areas to consider when deciding if an AI tool is the right one. Here are some practical and ethical considerations when assessing whether an AI tool is good enough for the job you need it to do.

 

 

Make a procurement checklist of current needs

 

Just because a tool does something impressive doesn’t mean to say that you actually need it. Create a checklist of the specific issues or sticking points you want an AI tool to address. What are the problems you want it to solve? Put the list in order of priority and don’t be swayed away from it.

 

For example, if you’re considering a podcast editing AI tool which does the task brilliantly but will only be used twice a year, then it might not be worth the cost, or time needed to train people. It can feel like there’s pressure to move fast on AI products, but always put strategy before speed.

 

 

Be a savvy consumer

 

Don’t be sucked in just because a tool is labelled “AI-powered”. AI washing, when companies make false claims about AI to promote their products, is tech’s equivalent to greenwashing. Do your research before buying a product, look closely at reviews, interrogate any claims, and ask for answers in plain English. If you feel out of your depth, try and find someone with expertise who can support you – maybe through mentoring or pro bono support.

 

Sense about Science have developed a helpful framework for the Responsible Handover of AI tools, which includes a flow chart of specific questions to ask a developer or vendor. These are wide-ranging and include:

• Where has the data come from?
• Is it likely to be biased?
• Is there other software or equipment the tool needs to interact with?
• How do you measure how accurate the tool is and is it being monitored?
• What are the skill requirements?

 

The framework can help you capture all the necessary information about a tool and the data that underpins it, so you can make a confident choice.

 

 

Ask for a free trial

 

If possible, use a free trial before committing to any fees. For example, Localgiving’s AI bid grant writer tool offers a free taster to see whether it’s a good match for your organisation. When considering fees, factor in any hidden costs, which might include training and technical support.

 

 

Don’t rely on demos

Ideally, test a tool using your own examples, not samples that have been chosen to give the best impression of it. Testing it against real cases will give a much clearer picture of whether it’s the right tool. Try to involve the people who would be using it as part of their role.

 

 

Work out how many hours you can save

 

If you’re trialling a tool, or working out whether to renew a contract, find out how quickly AI can achieve a useable output, and work out if that’s significantly different to how long it would take you without it.

 

Prioritise quality over quantity, recognising that not all AI outputs are up to scratch. For example, while your tool might create an image in a split second, if you’re then spending an hour adjusting the image until it’s usable, you’re not saving time. It might be quicker to do the whole job yourself.

 

Test the tool with a number of users and take an average. It’s likely that you’ll all be using it in slightly different ways, with different levels of expertise.

 

 

Check if it works with your existing systems and your team’s skillset

 

A product might be great, but it needs to work for you, your existing systems, and your team. Ask questions such as: How will it embed into what you already use? Will your team need significant training to be able to use it effectively? How will it actually be used day-to-day and will it just add another process?  

 

 

Consider any potential risks

 

Before buying or rolling out a new AI product, it’s important to carry out an AI risk assessment. Areas of potential risk include, data breaches, data bias, hallucinations, misinformation, the environmental impact, ethical misalignment, and reputational risks.

 

 

Do your research

 

There are lots of helpful resources available to support you in assessing AI tools. Take your time to research products carefully, rather than adopting them on impulse. It’ll be worth it in the long-run.

• The Charity Digital Code of Practice 2025, including the section on risk
• Charity Excellence Framework’s Data Protection Checklist
• Sense about Science’s Responsible Handover of AI Framework
• NCVO’s plan for choosing new software or digital tools