Grimaud Valat, lawyer, partner, and expert in digital law and personal data at DTMV Avocats reflects on his contribution to the report "Data Altruisme: Une initiative Européenne, les données au service de l'intérêt général," submitted by the Human Technology Foundation to Cédric O., Minister of State for the Digital Sector, on February 22, 2022. This report is the product of close collaboration between various institutional, private and academic players.
1. What is the context of the report? What is the reasoning behind the Foundation's discussion?
As soon as topics — turned real buzz words — like big data, AI and the exploitation of personal data come up, the conversation adopts a particularly negative tone.
Several recent scandals immediately come to mind, like Cambridge Analytica and cyber attacks on hospitals, as well as the lucrative exploitation of data by tech giants.
The potential of data exploitation, which can be done extremely ethically, is limited by such perspectives. It’s important to remember that artificial intelligence (AI) systems and algorithms that are fed by big data only do what humans created them to do, like all technology. The power of these technologies can also benefit great causes. A compelling example is the incredible work carried out by Imagia in Québec on detecting and diagnosing cancer through AI systems driven by medical imaging data. Developments of such scale that serve the public interest are only possible through mass processing of health data. The stigma surrounding AI systems must be eradicated before we can understand that they only have to be handled carefully.
This is one of our goals — to reverse this tendency, and to demonstrate just how much the technical advancements made in the area of data processing can benefit humanity, not harm it.
Clearly, the European Commission understands this, and in a somewhat forward-thinking way, in my opinion. As part of its digital strategy, the Digital Decade 2020-2030, the Commission proposed in the Data Governance Act a new legal concept that aligns altruism, public interest and data exploitation: data altruism. The goal is to initiate a social movement that makes data sharing for the public good a social norm. As such, the Foundation had to seize the opportunity to explore the concept and make it a reality, hence, the purpose of the report “Data Altruisme: Une initiative européenne, les données au service de l'intérêt général.”
The first recommendation of the report defines the perimeter of public interest as an essential dimension of trust : a perimeter that is too broad would be detrimental to trust, and a narrow definition would not take into account future uses. A balance must be found.
In the report, the Foundation includes keys to allow future data altruism players to execute their initiatives. While we must obviously believe in people’s power and their desire to want to do good, we must also know how to guide them a little. Recommendation 19 suggests the creation of a certification mark (label) for altruistic actors. The recommendation 20 also aims to create criterion in calls for tenders based on commitment to data altruism.
Finally, the report proposes concrete measures to encourage actors to share their data. For instance, recommendation 21 states the implementation of tax incentives for contributors and recommendation 26 proposes to define financial terms of access to data by categories of users.
2. The report presents various use cases: health, personal data, mobility, etc. According to you, which would be the most relevant to test the data altruism approach?
Each of the use cases that we have studied have been rather significant, whether they led to a practical result or not. In this regard, we've seen many players widely implement and rely on AI systems and the data that drive them to serve the public interest, which is very encouraging. Nevertheless, how the application of such systems across industries should be prioritized remains a matter of personal sensitivity.
For example, I think that they should be prioritized in the health sector for a number of several reasons. The people who took part of our work highlighted how data processing enhances diagnoses, treatment personalization and other factors that not only improve the chances of a patient's recovery, but that also optimizes patient care, which is crucial in the hospital sector that is already subject to significant stress.
Environmental protection is equally as highly strategic, in my opinion. Global warming is clearly and unfortunately a critical issue of this century. Taking responsibility for it is multifactorial and requires a close analysis of its causes and all other related data. For example, reducing air pollution in cities requires traffic data, like routes, means of transportation and temporal data, to be analyzed in order to redraw traffic plans, calculate carbon footprints, etc. We now have the technical ability to harvest and process high volumes of data, which could be game-changing.
The power of this technology could be used for the common good!
3. What are the main challenges to achieve efficient and ethical data altruism? What are the next steps?
The main challenge posed to data altruism is to regain the trust that has been sorely tested, for the reasons I mentioned previously.
The first level of trust involves ethics. The trust of individuals will be gained by helping them overcome their reflexive comparison between data sharing and privacy breaches. The trust of companies will be built by securing the data shared, by demonstrating that data sharing as part of data altruism will not lead to leaks to competitors and thereby to a loss of market advantage.
This challenge is paradoxical: the same people who share their personal data on social media are hesitant to share the same data under other circumstances, even if it is in the public interest. Data altruism organizations will have to assume their places in this ecosystem and be perceived as true trusted third parties.
The second challenge is to structure such organizations both technically and economically. The reliability and longevity of organizations responsible for data collection and management are intrinsic to fostering trust.
We now have to wait for the first data altruism organizations to be founded. Players, like the Human Technology Foundation, may be required to accompany such organizations in implementing what is necessary to rebuild this trust and, thus, to initiate a broad data sharing movement to the benefit of all.
* The Data Governance Act is a proposed legislation by the European Commission that aims to create a framework to boost data sharing with the goal of achieving data altruism.