Transparent AI for
charity needs
Bridgit uses AI to help charity teams build and set up their data integrations faster. AI is always optional, and we think you should know exactly how it works: what data is involved, if any, how it's protected and what you can stand behind with your trustees and supporters.
No buzzwords. No vague reassurances. Just straight answers.
- Bridgit works fully without AI. AI features are only used when your team chooses to use them
- Our AI builders write rules and logic without ever reading your records
- Only the optional Setup Wizard looks at samples of your data
- Your data is never used to train AI models
- Anything sent to Anthropic is deleted within 30 days
- Our full DPA and subprocessor list are published on our website
Three levels of AI use, all optional
Level 1: No AI (the default)
By default, none of your data goes anywhere near AI. You can set up and run every bridge in Bridgit without using an AI feature at all.
Level 2: AI Transformation Builder and AI Filter Builder
You describe the rules you need in plain English and the AI turns them into a precise set of instructions, handling the kind of complex conditions that would otherwise take hours to build by hand. That is where the AI's job ends. From then on, Bridgit's own (non-AI, traditional code) system carries out those instructions exactly as written, treating every record the same way every time, so no AI is involved in deciding what happens to your data. The AI only sees what you type into the chat and never reads your records, so we advise against pasting personal data into it.
Level 3: AI Setup Wizard (beta)
The Setup Wizard goes a step further. To propose how your data should be mapped and transformed, in line with how your organisation has handled it historically, it looks at samples of real records from your platforms and CRM. These may include donor names, email addresses or donation details, and they pass through Anthropic's Claude API to generate the proposal.
How Anthropic handles your data if you use the Setup Wizard
Where any data does reach Anthropic, it is used only to generate the response to your request. Under Anthropic's standard API terms, your data is never used to train AI models and is automatically deleted within 30 days. Bridgit never shares your data with other customers or uses it for any purpose beyond setting up and running your bridges.
Anthropic is listed as a subprocessor in our Data Processing Agreement, which sets out exactly what each AI feature sends.
AI suggests. Your team decides.
Whether it's a rule written by one of our builders or a mapping proposed by the Setup Wizard, nothing the AI produces is put to work until someone on your team chooses to use it. Once a rule is in place it runs the same way every time. The AI isn't making decisions about your data in the background.
Environmental impact
AI inference uses energy and water. We think pretending otherwise, or saying nothing at all, would be the wrong approach for a mission-driven organisation that serves nonprofits.
Here's where we are honestly: we're in the early stages of measuring our AI compute footprint. We don't yet have the numbers to publish, and we'd rather say that clearly than share estimates we can't yet stand behind.
What we can say is this: we use AI purposefully. Every AI feature in Bridgit exists because it saves real time for real people in the charity sector, not because AI is fashionable. We are actively working to understand our usage and its environmental cost, and we'll publish that data here as soon as we can do so accurately.
If your organisation has specific ESG reporting requirements, please get in touch. We're happy to share what we know and work through it with you.
Frequently asked questions
Straight answers to the questions charities usually ask us.
Talk to us
If you have questions about how Bridgit handles your data, or you'd like to discuss our AI approach before signing up, we'd genuinely love to hear from you. This stuff matters, and we'd rather have the conversation than have you wonder.