It's time to control health tech

Christopher Crecelius

Apple recently announced a breakthrough: Its newest AI model was trained on over 2.5 billion hours of movement and biometric data, collected through the Apple Heart and Movement Study.

The study included 160,000 participants who voluntarily opted to share their Apple Watch and iPhone data for research purposes. The models developed from this data now underpin tools capable of assessing fitness levels, flagging pre-diagnostic symptoms and even detecting pregnancy by using existing sensors, without the need for new hardware.

Although this could be a positive step towards digital health innovation, it also reveals something far more important: the immense power and value of large-scale, real-world biometric data. With enough granularity and time, health AI can infer patterns that once may have required clinical intervention or specialised diagnostic equipment.

However, there is one question big-tech companies are consistently avoiding. If your body is helping train the next generation of AI and billion-dollar money-making machines, who actually benefits?

Ownership and agency

As someone who has spent the past two decades building secure health data systems, often for the very institutions now deploying these tools, I believe this is the central ethical challenge of our era — not the progression of AI as a tool, but the ownership of the data it collects.

Wearables, fitness trackers and health apps can monitor every aspect of our biology, from heart rates and blood oxygen to sleep cycles and stress patterns. The marketing narrative of this data collection that we are sold is that we will “gain empowerment” and “insight.” However, I believe that the reality is much less attractive and more extractive opposed to liberating.

We all know one hard truth. Data typically flows in a single direction, and that is upwards into corporate systems. It fuels product development and predictive models, and generates commercial value. Simultaneously, the users who are producing this information are left with dashboards and passive, inadequate summaries. In research studies such as the Apple example, even these participants are likely to forfeit control over what happens next. Although they consented to data being used, they have no say in how the data is reused for future applications, and almost definitely no share in the value it will have created for Apple.

Defenders of the current system point to consent agreements and privacy policies. But in practice, consent is less a contract than a formality. Few users even read the terms and conditions as they passively press “I agree,” let alone understand what they’ve agreed to. Fewer still can revoke access or have a meaningful influence on how their data is used.

There’s also an even more profound imbalance at play — when individuals must choose between accessing modern health tools and surrendering their autonomy over personal data, it’s not a fair exchange but a coerced one. Very few apps allow unrestricted use without the user consenting to their data being accessed.

These systems are already being used to shape insurance pricing, clinical recommendations and mental health interventions. Suppose these tools are not built to benefit the user and broader society? In that case, it’s fair to say the organizations using the data have ulterior motives.

Regaining control

A growing movement is advocating for data sovereignty. The idea is that individuals should have genuine, enforceable rights over the data generated by their bodies. This vision goes far beyond privacy. It involves active access, transparency, profit and revocability. It calls for people to decide when and how their data is used, and to be able to withdraw that consent entirely.

The ability to participate in the value created, including when personal data is used to train commercial AI models or achieve medical breakthroughs, is an attractive solution.

Emerging decentralized tools and infrastructure make this model increasingly possible. It offers ways to secure, anonymize and share sensitive data without forfeiting ownership. In the modern era, it should be about designing systems that view people as partners in innovation, rather than passive data points.

Too often, we measure progress in terms of compliance: Is the system GDPR-compliant, is the data encrypted, is it stored in the correct jurisdiction? While these questions do matter, they are not enough. Compliance is a floor. What we need is a ceiling, a higher standard built around justice, agency and dignity.

Health data reflects the lived human experience of suffering, healing, stress, joy and vulnerability, and should be viewed as having much more value than just data to support machine learning. To capture and commercialize that data without meaningful participation from the individuals behind it is to strip away its context and humanity.

We stand at a turning point. The next decade of health technology will be defined not just by what it can do, but by whom it serves. We can continue down a path where biometric data is extracted and monetized by a small group of giant institutions, or we can build systems that foster intentional participation, share power and include not only the corporations but also the individuals involved in the rewards of discovery.

It all starts with the simple premise that if your body is generating the data, you deserve a stake in what it becomes.

Christopher Crecelius is the co-founder of AxonDAO, a decentralized science collective that gives individuals control over their health data, enables ethical AI insights, and funds breakthrough research, all governed by the community.