Donors are increasingly using ChatGPT, Claude and others to discover causes and help them decide where they should donate their money. Your donor can now discover your organization, understand your work, and complete a gift directly inside the AI conversation.
The AI landscape is ever-evolving, and after years of leveraging AI to help nonprofits raise money, we have turned our learnings into a powerful new tool called Agentic Giving so nonprofits can participate and benefit from giving through AI channels.
Today, Fundraise Up customers can turn Agentic Giving on in their Fundraise Up dashboard, just in time for Giving Season.
Agentic Giving from Fundraise Up makes sure that when an AI assistant acts on a donor's behalf, the gift still flows through your organization's official, authorized giving route.
Understandably, some nonprofit organizations leveraging AI still have concerns about the safety and security of AI — especially when it comes to financial information and donor fraud.
So let’s talk about safety, and how we’re protecting your organization, your finances, and your donors throughout the Agentic Giving process.
Is the donor in control of the Agentic Giving transaction?
- Yes. A human always confirms the gift.
- No donation completes without the donor explicitly confirming the nonprofit, campaign, amount, currency, frequency, payment method, and consent.
- There's no scenario where an AI system independently decides to give on someone's behalf.
Is a donor’s payment information safe?
- Yes. AI does not have reach beyond the chat session, and is not used to collect credit card information.
- Payment runs exclusively through your existing hosted checkout or through tokenized, wallet-based payment methods — the same kind of technology behind Apple Pay.
- No raw payment credentials are visible to an AI assistant.
- Your agent profile contains no donor or financial data. It publishes only what's already public: your organization's identity and your active campaigns.
- Supporter records, totals raised, and payment details are explicitly excluded.
- We don't receive anything from the conversation beyond what the AI agent chooses to send us through our own connection to it. On our side, we keep the standard donation record, plus — where Stripe makes it available — the donor's first name, last name, email address, and phone number.
Will Agentic Giving make changes to our standard donation process?
- No. Nothing about your existing checkout changes. Whether a gift starts in a conversation with an AI assistant or on your donation page, it runs through the same infrastructure, with the same fraud protection, the same compliance checks, and the same receipts, consent handling, and CRM sync your team already relies on.
What if the AI gets something wrong, like the amount, the campaign, or recurring schedule? Who's responsible, and can it be undone?
- Here’s what the Fundraise Up team can audit: every request an agent makes to our system is logged, and every donation is marked as agent-initiated and traceable back to the exact request that created it.
- As of today, Agentic Giving-completed gifts are one-time only. The donor approves the exact amount in their own digital wallet, and the agent can't change anything about the donation afterward without asking the donor again. If something's still wrong, it's fixed the same way any donation is — a standard refund through your dashboard.
Does Agentic Giving create new fraud risk?
- No new door opens for classic payment fraud, because the agent never touches card numbers. The donor pays and confirms the amount inside their own wallet. An agent can't charge more than was approved, can't go outside what a campaign allows, and can't charge the same gift twice.
- Each channel is protected by controls appropriate to how it works. For agentic checkout that means the donor’s personal confirmation in their wallet, strict limits on what the agent can do, request rate limits, and Stripe’s fraud protection and dispute handling on every donation.
Is Agentic Giving secure? Where is data stored, encrypted, and retained?
- Yes. Agentic Giving is covered by the same Fundraise Up is SOC 2 Type II, PCI DSS, and ISO 27001 certifications you already trust us for.
- From a data-handling and storage standpoint, an agent-initiated gift is treated exactly like any other donation moving through your existing checkout — same infrastructure, same storage, just a new source of donation.
Is publishing the agent profile itself a security risk?
- No. It only ever contains information that's already public, plus your active campaign details, and campaign details only appear in it after you've reviewed and accepted the terms. We control the structure of the profile, but the content comes from what the organization maintains in our dashboard, under the existing permissions.
Does Agentic Giving recommend some nonprofits over others?
- No. We're not building a charity marketplace, and we never rank nonprofits or make claims about which cause is "best."
- Your organization keeps the donor relationship. We're the infrastructure that makes sure that relationship survives the shift already underway.
Can we turn Agentic Giving off?
- Yes. You're always in control. Agentic Giving is off by default.
- An administrator has to review and enable it, and it can be turned off at any time.
The safety of your organization, your donors, and your finances are our top priority. As Agentic Giving grows, we will continue to innovate to bring you the best possible outcomes that will help you scale and grow to meet the needs of the people and communities you serve.
If you have any questions, please don’t hesitate to reach out. And if you’d like to see more technical information about Agentic Giving, please refer to this guide.