How We Prioritize Donor Data Privacy Using AI

We get asked a lot: is your data secure? Is it private?

We want to be crystal clear: we do everything we can to protect you, your donors, and your data.

In some contexts, that may mean security practices to safeguard against attack or malicious actors.

But in the context of AI, it’s about making sure that your data remains yours, and is not used to feed into an algorithm somewhere.

We're excited to shed light on our approach to data privacy, outlining how we employ our "Anonymous In, Personalized Out" process to ensure your privacy while delivering tailored AI-powered experiences.

‍

Understanding PII and the "Anonymous In, Personalized Out" Approach

PII, or Personally Identifiable Information, refers to data that can be used to identify an individual.

In privacy circles, PII is the compass guiding the fine line between personalization and confidentiality. PII includes things like names, addresses, phone numbers, and more.

In our discussions around privacy, one clear value held throughout the company is that simply complying with data protection regulations is not enough. Rather, we want to extend our privacy practices in far deeper, more substantive ways.

The "Anonymous In, Personalized Out" approach is the cornerstone of our data privacy strategy. This methodology involves removing any PII from the data before it's fed into our AI models. This means that any information that could identify a specific person, such as names and addresses, is carefully stripped away. What's left is the essence of the data — the meaningful context and patterns that drive the AI's predictions and insights — with none of the identifiable attributes.

Balancing Personalization and Privacy: An Example

"Anonymous In, Personalized Out" is easiest to understand through an example. Meet Robert Hernandez, a fictional donor.

Robert’s data in the CRM includes details about his financial goals, spending habits, and life circumstances.

1. Anonymizing Robert's Data: We begin by removing any PII from Robert’s data. His name, address, and other personally identifiable details are meticulously scrubbed. However, important contextual information about his life remains intact. For instance, while his wife's name is removed, the fact that he has a wife is retained through a placeholder.
2. AI Model Processing: The anonymized data is then fed into our AI model. The model leverages this contextual information to generate accurate insights and predictions, all while remaining oblivious to the individual's true identity.
3. Personalized Outputs: The AI generates outputs tailored to Robert’s financial situation and goals. These outputs are rich with meaningful advice and recommendations, ensuring that Robert receives a personalized experience without jeopardizing his privacy.

Striking the Balance: Re-Personalizing the Output

Here's where the magic happens. We understand that personalization enhances the user experience. To achieve this, we employ advanced software techniques that re-personalize the output in a privacy-conscious manner. This process involves taking the anonymized outputs and carefully applying algorithms to re-introduce personalized elements.

For Robert, this means that the advice he receives will reflect his unique circumstances. The AI-generated insights may suggest strategies that align with his goals, family situation, and spending habits. This re-personalization is achieved without exposing any sensitive information to AI models or violating his privacy.

Our Commitment to Donor Privacy

Data privacy is an integral part of our identity. The "Anonymous In, Personalized Out" process typifies our dedication to delivering powerful AI-driven solutions while respecting data privacy. We continually invest in cutting-edge privacy-preserving technologies and stay updated on the latest developments in data protection to ensure that your trust in us is well-placed.

As we move forward in this AI-driven world, rest assured that we will always prioritize your data privacy and security, while never compromising on the quality and personalization of your experience.

Remember, it's not just AI—it's AI with integrity.