Privacy-enhancing technologies (PETs) hold considerable promise for the public good. I learned this over a decade ago when I started building them into scrappy, high-stakes data collaborations where getting it wrong wasn’t an option. Here’s what I’ve learned since then: PETs are great tools for specific problems, not for every problem.

Privacy-enhancing technologies (PETs): Technologies that minimize an information system’s possession of personally identifiable information (PII) without compromising functionality. 

This nuance was front and center this past summer at Georgetown University’s McCourt School of Public Policy, where I joined the Summer Institute on PETs for Education Data. It’s always great to be back at Georgetown, and I’m grateful that McCourt and the Massive Data Institute (MDI) continue to center privacy-enhancing technologies as a core component of policy training. Over three days, technologists, policymakers, and researchers explored how PETs could improve statewide longitudinal data systems, strengthen research capacity, and protect individual privacy.

As we looked at tools and techniques, we returned to a deeper question I’ve been sitting with for a while—one that cuts across toolkits, technologies, and frameworks:

How do we build civic and institutional cultures where PETs can best succeed?

Data Sharing is Civic Work

Sharing data sits at the intersection of technology, trust, community memory, and governance. At that intersection lies both risk and opportunity.

Education data, for example, is incredibly powerful. It tells the evolving story of a learner’s journey, a dataset that reflects an individual’s interactions with various systems over more than a decade. Used wisely, it helps identify which programs are working, where support is needed most, where students are falling through the gaps, and how to address systemic inequality. Such data often also involves sensitive variables, such as housing instability, disciplinary records, or familial circumstances, which means misuse or careless handling can stigmatize learners, reinforce bias, or expose families to unintended surveillance and discrimination.

PETs add value when integrated into strategies rooted in trust, transparency, and shared civic purpose. 

Right Tool, Right Job, Right Context

Emerging data-privacy technologies sometimes get positioned as universal solutions to data-sharing challenges. Some technologies do lean toward being more general-purpose; PETs, however, are not generalized, one-size-fits-all solutions. Their true value comes from being applied in the right contexts. PETs fit best where the stakes of sharing sensitive information are the highest:

Regulatory sensitivity. Because of compliance and legal restrictions, some data cannot be pooled in one location. For example, there may be situations where health and education data intersect in a manner that PETs enable collaboration through helping meet compliance demands.

Social sensitivity. In some contexts, such as domestic violence services or legal aid organizations, the consequences of a data breach may be too severe to risk conventional sharing. PETs help protect participants in situations where exposure could cause serious harm.

Political sensitivity. Cross-agency or cross-jurisdictional collaborations among Tribal, state, and federal partners may require new safeguards. In such cases, cultural or structural mistrust makes centralizing data difficult and, sometimes, impossible. PETs enable collaboration without forcing sensitive data into a single, politically fraught location. 

Knowing the importance of context for data-sharing approaches, we see PETs not as “magic bullets,” but as carefully applied solutions that respect the regulatory, social, and political contexts in which sensitive data exists.

Designing with Accountability in Mind

What about implementation? Will it be too complicated? These are common causes of hesitation around PETs, and the concerns are valid. PETs aren’t turnkey. They require advanced technical skill, thoughtful design, and necessary cross-sector coordination. They also introduce governance questions that some organizations aren’t yet equipped to answer. Implementation demands careful planning and tight coordination.

But that’s a feature, not a flaw.

At the Summer Institute, this distinction came into focus. Complexity isn’t a reason to abandon PETs. If anything, PETs can offer a higher standard for how to approach all emerging technologies. Unlike current trends in generative AI—where systems are rapidly deployed with questionable guardrails—PETs’ narrow purpose demands more intentionality. They force three essential questions: 

• What exactly are we protecting?
• How do we protect it? 
• Who decides whether that protection is effective?

If implemented correctly, PETs bake accountability to complex governance questions into the process. They invite us to consider governance, consent, and legitimacy before data is ever touched. This approach drives Asemio’s design processes, ensuring humans stay at the heart of every data project. 

And here’s where the real opportunity lies: PETs push us toward design that’s anticipatory, not reactive. They invite us to raise the bar for ethics, public trust, and technical excellence in how we handle data.

What We’re Building in Oklahoma: Privacy-Enhancing Technology in Context

Through the Tulsa Tech Hub grant, we’re supporting one of our state’s more ambitious efforts to safely connect education and workforce data. The goal is clear: Align training with economic opportunity, invest wisely, and serve residents without compromising their privacy.

PETs are part of that picture, but not its entirety. They enable linkages across education, labor, and community organizations when sharing raw data isn’t feasible, which is a pressing challenge in Tulsa and many cities across the nation. But success depends just as much on governance frameworks, shared ethics, and a culture that values both innovation and caution.

This is also why implementation guidance and standards matter. State and local leaders need improved best practices and more accessible guidance for agencies and practitioners. On this front, data experts at MDI recently released a valuable how-to guide on safely implementing one type of PET, synthetic data, with more guides planned for the future.

Preparing for What Comes Next

Toward the end of the Summer Institute, someone asked me a deceptively simple question: How do we prepare people for a future that never stops changing? My answer wasn’t about technology, because a useful answer always requires more than technology. Instead, I shared my belief that we can prepare for an accelerated pace of change by building two core capacities:

The first is adaptability. When people exist in a culture of continual learning, they’re better equipped to handle shifting tools, standards, and risks. Because toolsets change so rapidly, learning how to adapt quickly provides more value than focusing on any one tool.

The second is empathy. Civic systems designed without empathy inevitably miss the mark, no matter how advanced the technology. That’s because these datasets represent lives, families, and communities. Radical empathy ensures we never lose sight of the real people who are the heartbeat of every system.

If we can cultivate adaptability and empathy simultaneously, we’ll do more than implement PETs effectively. We’ll create institutions that are resilient, human centered, and trustworthy. PETs can play a role in that evolution, but they aren’t the end goal.

The hardest problems to solve are rarely technical. They’re cultural.

Want to help document best practices for privacy-enhancing technologies? Email info@asemio.com to start the conversation.

​