Good Intent and No Surprises

Jim Fruchterman | August 12, 2025

Note: In the following essay we reference the eight Better Deal for Data™ commitments originally proposed in our April 2024 white paper. In December 2025, we released the BD4D™ Commitments as a set of seven refined commitments.

Introduction

The Better Deal for Data (BD4D) is an approach to data governance that relies on good intent. The first of the eight BD4D Commitments requires that the organizations adopting the Commitments use the data they collect for the benefit of the people being served (“You and Your community” in the language of the Commitment). The entire premise of the Better Deal revolves around building and deserving trust, which is generated by backing up good intent with positive actions.

One test for trustworthiness is avoiding unpleasant surprises. If the community you serve would be negatively surprised by how you are handling their data, your approach is probably not a fit for the Better Deal for Data. You need to intend to benefit the community, but that must align with the community’s understanding of what’s beneficial to its members. This consideration underscores the importance of context: one data use case that would be fine with one community might be anathema in another community.

This tension is what leads us to the Major Question in this paper: what does it mean to have ‘No Surprises’ under the Better Deal for Data?

We think there are three priorities that need to be top of mind when making the decisions guiding a program’s data collection and use policies. These priorities are key to avoiding surprises, and will be the focus of this paper. The first priority is going to some length to ensure that the benefits primarily flow to the community being served, and that these benefits are easily understood by that community. The second priority is to ensure that activities align with community expectations about how their data will be used, and if shared, how that sharing is ethical and beneficial. The third priority we will discuss is the need to be transparent about an organization’s actions, partners, and vendors. We will also touch upon the issue of consent, and how it fits under the Better Deal.

This working paper is one of a series we are publishing on major questions we are being asked about implementing the Better Deal for Data. It is not intended to be the final word on where the Better Deal for Data standard ends up on these questions. Instead, these Major Questions essays are encouraging us to explore each specific question, and think through our initial answers. We are hoping to engage in more extensive debates over any controversial points.

Benefits

Since the primary purpose of data collection and use under Commitment One of the Better Deal for Data is to benefit individuals, communities, humanity, and the planet, you should be able to explain in straightforward language what those benefits are. The communities you serve should not find it difficult to understand how and why you are operating in their best interests. If there are possible negative consequences to your program, it should be clear that the positive impacts considerably outweigh the negatives.

The second half of Commitment One is that the activity is not being conducted for private gain or profit. If your organization is structured as a charity, this is theoretically baked into your charter, but a charter is not a guarantee of good behavior.

If benefits only flow to the communities being served, say in the provision of free services by a completely volunteer work force, this Commitment is not hard to meet. But, most programs need staff and philanthropic support to deliver their programs, including for data-related activities. Fund raising, revenue, and staff salaries should be in line with reasonable practices in the field. A good first step to align with this Commitment is to minimize data collection, retention, and sharing in the first place. For certain contexts, this might be the best course of action. For example, organizations working with communities that are targets of active discrimination, exploitation, or harms often choose not to collect data which could subject their members to risks if that data were exposed, seized, or stolen.

However appropriate data minimization may be in certain contexts, the Better Deal for Data overall is based on the belief that ethical data sharing can be the source of major new benefits for humanity and the planet. The for-profit industry has demonstrated the power of large data sets to create wealth for the few; we think that the nonprofit use and sharing of data can benefit the many. Improving the effectiveness of nonprofit programs, as well as advancing the progress of scientific research, are benefits that we believe most communities would support and would not see as a source of unpleasant surprises.

Expectations

Beyond delivering tangible social benefits, Better Deal for Data adopters need to ensure that the conduct of their data activities matches up with community expectations. If most members of the community who read the BD4D Commitments would be unpleasantly surprised if they knew what was actually going on with their data, you are not following the Better Deal for Data! Let’s explore the key elements of living up to expectations.

Make sure that data collection and use matches up with the intended benefits. If your detailed activities might be controversial in the community you serve, you have a special obligation to be transparent with your plans (or reconsider them). Of course, being more transparent (as discussed below) and engaging the community regularly are proactive ways to avoid surprises. If the community reaction to a change in data policies or increased awareness of current data policies is broad or strident, this is an indicator that you have failed to avoid surprises, and identifies a critical need to re-evaluate your project/policies.

You should be aware of power imbalances both inside the communities you serve, and between the data collecting organization and those communities, to ensure that the data use actually benefits the majority of the community, not just the loudest or most powerful actors, and that data collection is not implicitly coerced.

Projects that are likely to be controversial should consider engaging in more extensive community engagement activities. This might include the sort of effort envisioned in the Social License for Data concept proposed by The Data Tank, where the project convenes a representative subset of the community to review the project and its data use plans. In some cases, a formal informed consent process may be required (by an Institutional Review Board, or by law), or be advisable even if not required.

If you are sharing confidential or sensitive data with aligned organizations, the group you are sharing the data with should both commit to the BD4D Commitments (in writing), and be seen by the community as a trustworthy partner. Adopting the BD4D Commitments is not a way to justify sharing data with groups the community does not trust. It is also incumbent on a Better Deal for Data adopter to actively avoid sharing sensitive data with vendors and tech companies for their disclosure or reuse in commercial products. As discussed in AI and the Better Deal for Data, this means ensuring that AI companies are not training on confidential or sensitive data.

Organizations adopting the Better Deal operate in the real world. Not everyone in the communities you serve will agree with all your decisions, even with the best intent. The ability to opt out or delete one’s data is intended to address the needs of specific individuals. It is not intended as a substitute for actively putting the interests of the community first. But, it does allow community members to change their minds if some decide later they don’t like what’s happening with their data.

Transparency

If you make information about your data-related activities and policies easy to find and understand, you greatly lower the chance that someone in the community will be surprised. Now, many people served by the nonprofit sector trust the organizations helping them, and don’t want to know the details, or don’t have the background to understand technical or legal nuances. Even if you do not have a comprehensive informed consent process, you still need to make it straightforward for community members to learn about your activities which are using their data, and how those activities affect, or could affect, them.

Both your activities and associated transparency need to be aligned with the Better Deal principles: you can’t use transparency claims as a fig leaf for doing something that clearly violates the BD4D Commitments. As noted in the previous section, there may be projects where transparency is not sufficient to avoid negative surprises. In these cases a more active process may be needed, such as the Social License mentioned above, or following the Ada Lovelace Institute approach to participatory data stewardship.

Transparency also means being clear about what deletion of data upon request entails. It may be impossible to erase data which was not retained (it’s already been deleted), data which has been anonymized, or data which contributed to aggregated reports (e.g. how many people were helped in the previous year). It may be illegal to delete certain data, or limited data may need to be retained to handle future inquiries. In practice, data deletion requests take time to fully execute, because many tech platform providers hold backups for thirty or sixty days in case of a systems failure. Diligent data deletion efforts aligned with these practices will be compliant with the BD4D Commitments.

Note that data will often be shared with the tech platform providers supporting an organization’s data operations, who have agreed to not reuse the data. For example, the three main cloud services providers (Amazon Web Services, Microsoft Azure, and Google Cloud) treat the data of their paying customers confidentially. They contractually agree to not reuse this data, and violating that agreement would be against their economic interests. In such cases, this use of reliable tech platforms does not rise to being a major disclosure item in practice.

As a counterpoint, many nonprofits use “free” plans from technology vendors, and these vendors often appropriate data from their free plan users for commercial purposes. Organizations adopting the BD4D Commitments need to take care that sensitive data from the communities they serve is not being hijacked by their tech vendors. We further discuss this issue in our AI and the Better Deal for Data paper.

Consent

We recognize that most nonprofits cannot practically require explicit informed consent for each of their program’s activities. We also don’t believe that permissions buried in a 20-page Terms of Service are real consent, despite how prevalent this type of ‘notice & choice’ agreement is online. The Better Deal for Data generally offers a trustworthy alternative, by making strong commitments to shared social values and creating an implied permission to operate within the communities being served.

However, this is not universal. Informed consent is the norm for human subjects research, such as medical or behavioral experimentation. In some jurisdictions or in certain activities, it’s legally required. If it is otherwise required or the norm, then it is also required by the Better Deal, because to do otherwise would be a negative surprise.

Conclusion: What’s Missing?

The Major Questions papers are intended to explore the big issues we hear about from the many collaborators who have contributed their data use cases, their feedback, and their support to the Better Deal for Data. The foregoing distills our initial thinking about the concept of “No Surprises.” At this point, we would like to get even more feedback from the community, especially:

• What additional issues come to mind about this subject?
• What did we get wrong?
• What examples do you have of nonprofit data use which should be inside or outside the Better Deal for Data, or are simply puzzles to consider?

Delivering community benefits, meeting expectations, and operating transparently are central to avoiding surprises. Keeping these in mind while working with data greatly increases the real, and perceived, trustworthiness of social sector organizations using data to drive their programs.

We are looking forward to many new questions and ideas as we work together to craft a usable Better Deal for Data.