Responsible AI

If you're cautious about AI, that's a reasonable place to start from.

Here's exactly where we stand.

Human oversight stays with you.

AI supports decisions. It doesn't make them. Every recommendation, every output, every automated process is something your team reviews and controls — not something running unsupervised in the background.

Privacy and data security come first.

We won't put your data, your beneficiaries' data, or your funders' data through a tool without understanding exactly where it goes and how it's protected. If we're not confident in a tool's privacy position, we won't recommend it.

We tell you what AI can't do.

Overselling AI is how organisations end up disappointed, or worse, exposed. Part of our job is telling you when the honest answer is "AI isn't the right fit for this."

Governance matters as much as the technology.

Part of what we help build is the policy and oversight structure around AI use — who's responsible for what, how mistakes get caught, how decisions get reviewed — not just the tools themselves.

Where the boundary sits

AI prepares. Your people decide. That line doesn't move.

Where the boundary sits between AI and your team Two bands separated by a firm line. The upper band, "Your team decides", covers approving, deciding and being accountable. The lower band, "AI prepares the groundwork", covers drafting, summarising and spotting patterns. Three arrows rise from the lower band and stop at the line, showing that AI's work feeds up to your team but never crosses into deciding. Your team decides Approves Decides Accountable This line doesn't move AI prepares the groundwork Drafts Summarises Spots patterns

AI can draft the report, pull six months of data into something readable, and take the first pass at the paperwork. What it never does is sign it off. A board, not an algorithm, makes the call — and that stays true no matter how much of the groundwork moves underneath.

Two people talking one to one over an open notebook in a community office.

Talk to us about your organisation.

Technology should serve people, not the other way around — including knowing where AI shouldn't be used.