You are on page 1of 5

Commentary

Big Data & Society


July–December: 1–5
AI ethics and data governance in the © The Author(s) 2022
Article reuse guidelines:
geospatial domain of Digital Earth sagepub.com/journals-permissions
DOI: 10.1177/20539517221138767
journals.sagepub.com/home/bds

Marina Micheli1 , Caroline M Gevaert2 , Mary Carman3 ,


Max Craglia1, Emily Daemen4, Rania E Ibrahim5 ,
Alexander Kotsev1 , Zaffar Mohamed-Ghouse6, Sven Schade1 ,
Ingrid Schneider7, Lea A Shanley8 , Alessio Tartaro9
and Michele Vespe1

Abstract
Digital Earth applications provide a common ground for visualizing, simulating, and modeling real-world situations. The
potential of Digital Earth applications has increased significantly with the evolution of artificial intelligence systems and
the capacity to collect and process complex amounts of geospatial data. Yet, the widespread techno-optimism at the
root of Digital Earth must now confront concerns over high-risk artificial intelligence systems and power asymmetries
of a datafied society. In this commentary, we claim that not only can current debates about data governance and ethical
artificial intelligence inform development in the field of Digital Earth, but that the specificities of geospatial data, together
with the expectations surrounding Digital Earth applications, offer a fruitful lens through which to examine current
debates on data governance and artificial intelligence ethics. In particular, we argue that for the implementation of ethical
artificial intelligence and inclusive approaches to data governance, Digital Earth initiatives need to involve stakeholders and
communities at the local level and be sensitive to social, legal, cultural, and institutional contexts, including conflicts that
might arise within those contexts.

Keywords
Artificial intelligence ethics, data governance, digital earth, local context, geospatial domain, ethics guidelines

Introduction
Digital Earth (DE) can be defined as a multidimensional 1
Digital Economy Unit, European Commission Joint Research Centre,
representation of the planet capturing natural, social, and Ispra, Italy
cultural phenomena (Goodchild et al., 2012; ISDE, 2012). 2
Faculty of Geo-Information Science and Earth Observation, University of
Underpinned by both structured and unstructured, open or Twente, Enschede, the Netherlands
3
proprietary, spatial–temporal data, DE makes use of Faculty of Humanities, University of the Witwatersrand, Johannesburg,
South Africa
digital technologies to visualize, simulate, and model real- 4
Geonovum, Amersfoort, the Netherlands
world situations. The concept of DE, initially popularized 5
Scientific Publishing and Documentation, National Authority for Remote
by Gore (1998), has long been loaded with the Sensing and Space Sciences, Cairo, Egypt
6
techno-optimism that often surrounds new technologies. Department of Infrastructure Engineering, The University of Melbourne,
DE has even been envisaged as a platform to support Melbourne, Australia
7
Faculty for Mathematics, Informatics, and Natural Sciences, Department
national and international cooperation for global sustainable
of Computer Engineering, Universität Hamburg, Hamburg, Germany
development, economic growth, and well-being (ISDE, 8
The International Computer Science Institute (ICSI) in Berkeley, CA
2012). The optimistic vision of DE from 10 years ago, 9
Dipartimento di Scienze umanistiche, Università degli Studi di Sassari,
however, could not fully anticipate the impact of big data, Sassari, Italy
machine learning, and artificial intelligence (AI) systems
Corresponding author:
on the field.1 These are crucial for DE’s development due Marina Micheli, European Commission Joint Research Centre, Digital
to the high volumes of data coming from multiple sources Economy Unit, Via Fermi 2749, 21027 Ispra, VA, Italy.
—from remote sensing to crowdsourcing projects—and Email: marina.micheli@ec.europa.eu

Creative Commons Non Commercial CC BY-NC: This article is distributed under the terms of the Creative Commons Attribution-
NonCommercial 4.0 License (https://creativecommons.org/licenses/by-nc/4.0/) which permits non-commercial use, reproduction and
distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access page (https://us.
sagepub.com/en-us/nam/open-access-at-sage).
2 Big Data & Society

the need to integrate data to build complex multiscale and data that feed DE (Annoni et al., 2011; Goodchild, 2007).
multisector models. Issues at stake with citizens’ involvement (or lack thereof)
In this commentary, we argue that DE offers a fruitful in DE relate well to such debates on data governance
lens through which to address current debates on data gov- (Brovelli et al., 2020; Haklay, 2016).
ernance and the ethical use of AI. In our view, the many Expanding on the ideas in this introduction, in the
important issues being discussed about the ethics of AI second section we discuss the implications of AI ethics
are fundamentally about both the data and the algorithms for DE, in the third section those of data governance, and
used by AI. We propose that not only can current debates in the fourth section we conclude.
about data governance and ethical AI inform development
in the field of DE, but also the distinct characteristics of
DE (and of the geospatial information from which it is AI ethics, DE, and the local dimension
built) can also inform data governance and ethical AI Many guidelines have been released stating principles
debates more generally. and values for AI ethics. Globally, the UNESCO
There are many reasons why DE and the geospatial Recommendation on the Ethics of Artificial Intelligence
domain provide valuable insights for discussions on data (2021) is one of the most prominent, providing a “universal
governance and ethical AI. framework of values”. Recommendations like UNESCO’s
First, while the collection of geospatial data by private are not legally binding. Nevertheless, public and private
sector operators is becoming ubiquitous (Poom et al., organizations across the world are already voluntarily
2020), geospatial data about an individual’s location and holding themselves accountable to various AI ethics princi-
proximity to other features on a map can reveal privacy- ples and guidelines (Jobin et al., 2019).
sensitive information. Thus, the ethical handling of geo- Studies have attempted to identify convergences and
graphic data, referred to as GeoEthics within the GIScience divergences in the available ethical guidelines (Jobin
community (Goodchild, 2021), has an important role to et al., 2019; Vesnic-Alujevic et al., 2020). These compara-
play in general discussions on the ethics of data and AI. tive analyses reveal two main tendencies. On the one hand,
Second, DE involves both large volumes of spatial–tem- guidelines converge on a set of ethical principles, including
poral data from multiple sources—across different owner- transparency, justice, non-maleficence, responsibility,
ship and governance regimes, sectors, scales, and cultures accountability, privacy, safety, and trust. On the other
—and complex interactive models, AI methods, and techni- hand, guidelines lack indications on how to implement
ques.2 Instances of DE, like interactive digital twins of the these principles in the research, development, and applica-
Earth and its inhabitants, are capturing the interest of policy tion of AI systems. In other words, they identify the “what”
makers for their potential to provide tailored models for of AI ethics, but neglect the “how” (Morley et al., 2020).
policy, monitoring, and simulating possible futures Furthermore, the very meaning of these principles is
(Nativi et al., 2021). continuously renegotiated and adapted in local contexts.
Third, the vision and role of DE as a global platform for The diversity of local cultures and settings makes adapting
collaboration across governments and cultures, by being “umbrella principles” to a local context challenging. Giving
multilayered and embedded in different local contexts, local perspectives power to help frame and balance these
offers a nuanced case study to examine some of the global- priorities is critical (Carman and Rosman, 2020; Gevaert
local tensions that are arising in debates about AI ethics and et al., 2021). For example, a principle that “individuals
data governance. have a right to make decisions for themselves” (Floridi
Current policy and academic discussions on AI ethics and et al., 2018: 697) is prominent in many Western cultures.
data governance could also offer valuable insights for the Yet, other cultures across the world may promote the role
future of DE. Nations across the world such as China, the of the community in decision making (Chukwuneke et al.,
USA, India, Canada, and member states of the European 2014). Similarly, the AI ethics policy of the New South
Union have developed AI strategic plans, ethical guidelines, Wales Government in Australia has included “community
and recommendations for responsible AI (e.g. European benefit” as one of its five principles (NSW Government,
Commission, 2018; UNESCO, 2021; Wu et al., 2020). 2020). Local values related to culture, sovereignty, self-
Yet, there is little guidance on how to practically implement determination, sense of privacy, etc., have to be taken
these guidelines in DE applications and how to develop data into account if a more responsible development and appli-
governance structures that support these values. cation of emerging technologies is sought (Carroll et al.,
Furthermore, within the data governance field, scholars and 2019; Shanley et al., 2015).
policy makers are problematizing the unequal distribution An example of how such challenges might unfold in
of power and investigating how individuals can be empow- practice in DE is the way AI algorithms can be used to auto-
ered in their relation with their data (Micheli et al., 2020; matically generate maps from satellite and drone imagery
Sadowski et al., 2021). Citizens have a key role in DE, on a global scale. While this requires sharing the imagery
both as final users of applications and as co-generators of with the experts and teams developing the algorithms, a
Micheli et al. 3

project using drone imagery to map informal settlements, interest. For example, the National Authority for Remote
might involve local communities in identifying which Sensing and Space Sciences in Egypt established a
objects in the images could cause privacy concerns. In GeoPortal to share geospatial information and to collect
this way, possible strategies to address residents’ concept data from the public about private businesses nationwide
of privacy can be developed (e.g. blur orthoimagery of sen- (Ibrahim et al., 2021).
sitive objects) and the first step is made toward empowering The public value of geospatial information emerges even
citizens and building trust (Gevaert et al., 2018). At the same more during emergencies and natural disasters (Mulder
time, potential conflicts in values could arise between levels et al., 2016). For instance, during the COVID-19 crisis,
(e.g. local vs. regional) and even at a local level, where dif- policy makers relied on data collected by mobile devices
ferent parties may disagree and power asymmetries come to study the spread of the virus and track recovery (Poom
into play. While local instances of DE allow some custom- et al., 2020; Vespe et al., 2021). Citizens themselves con-
ization of data and deviation from global “standards”, their tributed through crowdsourcing projects such as the
practical implementation will be challenging due to such fric- Italian OpenStreetMap (OSM) community, which produced
tions. This illustrates why DE can be a good case study for AI mapping with a scale and granularity higher than datasets
ethics: it is a global but multilayered framework that “moves” collected by public sector administrations at that time
between global and local. (Minghini et al., 2022).
To be inclusive, however, data initiatives should account
for the perspectives of members of different sociodemo-
Inclusive data governance for DE graphics and communities and produce benefits also for
In datafied societies, substantial asymmetries exist between the less privileged. For instance, in the geospatial domain
those that collect and control data, and other actors with less top-down earth observation is often used to collect informa-
negotiating power, such as individuals, communities, local tion for urban planning. Yet, the inclusion of bottom-up per-
governments, and small businesses. In the DE field, former spectives in the data governance of such initiatives can
US Vice President Al Gore and others raised awareness improve the capacity to address societal concerns and
long ago about how the restriction of geospatial data to needs of communities underrepresented within an exclu-
the hands of governments and corporations could impede sively top-down approach. Such a bottom-up approach is
societal advancement (Goodchild et al., 2012). The notion illustrated, for instance, by a project on slum mapping in
of DE envisions citizens as key actors, both as data users Ghana, where scholars produced user-driven slum maps by
and data producers through crowdsourced and citizen/com- carrying out in situ observations and interviews with
munity science projects (Brovelli et al., 2020), but citizens’ experts from local institutions addressing the impact of
role in the governance of data is weak (Micheli et al., 2020; maps on slum dwellers (Owusu et al., 2021). Members of
Mulder et al., 2016). marginalized communities can be represented by boards
Data access, control, and use by a wide range of actors is and dedicated organizations. For instance, American Indian
a prerequisite for an inclusive approach to data (Carroll tribes rely on indigenous organizations that steward data to
et al., 2019; Shanley et al., 2015). Proposals for more inclu- protect their cultural knowledge and way of life (Carroll
sive data governance are emerging that could inspire the et al., 2019). Overall, grounding (geospatial) data govern-
governance of DE. These proposals advocate for the redis- ance in local contexts is part of what makes these initiatives
tribution of value generated from data across society; the more inclusive and sustainable (Mulder et al., 2016).
ability to exert authority (sovereignty) over data by indivi- Including local perspectives involves addressing conflicts.
duals and organizations; and the inclusion of less powerful Local responses might be bottom-up data resistance initia-
actors, such as marginalized communities, in data govern- tives, intended as local activities that resist top-down govern-
ance so they can have claims on the benefits derived from ance, such as location spoofing or counter-mapping (Dalton
data (Micheli et al., 2020; Mulder et al., 2016; Taylor, and Stallmann, 2018; Dalton and Thatcher, 2014).
2017). Understanding how such approaches are integrated Furthermore, there can be local groups with conflicting inter-
into the geospatial domain can provide guidance for the ests, such as claims to the same territory/resources, but with
future of DE. different cultural and organizational perspectives toward the
City governments, for instance, are experimenting with land and its data (Cullen, 2015; Kyem Kwaku, 2004). Local
the opportunities offered by geospatial data to tackle resilience strategies, dedicated spatial planning initiatives,
urban challenges and create public value, by providing, and creating a shared social aim could be means to
for example, citizen-centric public services and improving respond to intra-community conflicts (Haugh, 2021).
governance. Public value creation through geospatial data
at a local level also takes place when using digital twins
of cities to test policies before adoption. Partnerships Conclusions
between governments, science, and industry are crucial to In this commentary, we highlight how the tension between
develop collaborations and to share data for the public DE’s local embeddedness and global scope makes it a
4 Big Data & Society

compelling case for current debates on data governance and and society: Developing responsible and viable solutions for geo-
the ethical development and use of AI. Building on this, we spatial data” (with project No. 18091).
argue that the implementation of more inclusive data gov-
ernance approaches and ethical AI in the geospatial ORCID iDs
domain has a strong local dimension. Good practices high- Marina Micheli https://orcid.org/0000-0002-7004-6480
lighting community-based approaches should be collected Caroline M Gevaert https://orcid.org/0000-0002-3983-2459
and critically examined to understand how guidelines and Mary Carman https://orcid.org/0000-0002-8163-2100
principles can be applied in local contexts to ensure inclu- Rania E Ibrahim https://orcid.org/0000-0001-5116-5866
sive innovation. Careful consideration should be given to Alexander Kotsev https://orcid.org/0000-0003-0411-741X
understanding how principles and values underlying DE Sven Schade https://orcid.org/0000-0001-5677-5209
vary across geographic contexts, and to promote intercul- Lea A Shanley https://orcid.org/0000-0001-8449-4615
tural dialogue to inform future implementation of laws Alessio Tartaro https://orcid.org/0000-0002-0382-3083
Michele Vespe https://orcid.org/0000-0003-2309-7266
and guidelines. Nonetheless, conflicts should be expected
to occur, as balancing the values and interests of actors
and communities is challenging at all levels. A crucial chal- Notes
lenge for DE is thus not how to reproduce a standardized 1. AI can be defined as “any machine or algorithm that is capable
replica of the entire planet, but how to account for the of observing its environment, learning and, based on the
(many) localized needs, values, and ethics that constitute it, knowledge and expertise gained, taking intelligent action or
and to do so in a way that is sensitive to conflicts and proposing decisions”. Machine learning is a subfield of AI in
which a behavior is predicted from a number of input examples
power asymmetries.
(Craglia et al. 2018:23)
This commentary encourages research in DE to engage 2. Different data layers can serve as input to DE, both open and
with current debates on ethical AI and data governance, proprietary, from maps produced within the OSM community
and to collect and analyze cases that implement principles to information collected via remote sensing.
of ethical AI and more inclusive data governance in the geo-
spatial domain. In turn, as DE provides a case from which References
other fields can learn, we encourage establishing a stronger
Annoni A, Craglia M, Ehlers M, et al. (2011) A European perspec-
dialogue with disciplines outside the geospatial field.
tive on digital earth. International Journal of Digital Earth
4(4): 271–284.
Acknowledgments Brovelli MA, Ponti M, Schade S, et al. (2020) Citizen science in
support of digital earth. In: Manual of Digital Earth.
The content of this commentary is based on the outcome of a Singapore: Springer Singapore, pp. 593–622.
workshop on “Digital Governance and AI Ethics” organized by Carman M and Rosman B (2020) Applying a principle of explic-
the International Society of Digital Earth (ISDE) Working ability to AI research in Africa: Should we do it? Ethics and
Group 3 on Digital Earth Governance and Ethics, and on subse- Information Technology 23(2): 107–117.
quent discussions between the co-authors of this piece. More Carroll SR, Rodriguez-Lonebear D and Martinez A (2019)
information can be found here: http://www.digitalearth-isde.org/ Indigenous data governance: Strategies from United States
list-94-1.html. The Working Group is open to contributions native nations. Data Science Journal 18(1): 31.
from a wide range of disciplines and welcomes contributions Chukwuneke F, Umeora O, Maduabuchi J, et al. (2014) Global
and exchanges on the topics presented in this commentary. The bioethics and culture in a pluralistic world: How does culture
authors would like to thank the anonymous reviewers and influence bioethics in Africa? Annals of Medical and Health
the editors for their valuable comments that helped to improve Sciences Research 4(5): 672.
the manuscript. Craglia M, Annoni A., Benczur, et al. (2018) Artificial Intelligence
– A European Perspective. Luxembourg: Publications Office
of the European Union.
Declaration of conflicting interests
Cullen A (2015) Making sense of claims across institutional
The author(s) declared no potential conflicts of interest with divides: Critical PGIS and mapping customary land in
respect to the research, authorship, and/or publication of this Timor-Leste. Australian Geographer 46(4): 473–490.
article. Dalton CM and Stallmann T (2018) Counter-mapping data
science. The Canadian Geographer/Le Géographe Canadien
62(1): 93–101.
Funding Dalton C and Thatcher J (2014) What does a critical data studies
The author(s) disclosed receipt of the following financial support look like, and why do we care? Seven points for a critical
for the research and/or authorship of this article: The work of indi- approach to ‘big data’. Society and Space 29. Available at:
vidual authors may be (partially) funded by grants. For example, http://societyandspace.com/material/commentaries/craig-dalton-
the work of Caroline Gevaert is (partly) financed by the Dutch and-jim-thatcher-what-does-a-critical-data-studies-look-like-and-
Research Council (NWO) through the Talent Program Veni why-do-we-care-seven-points-for-a-critical-approach-to-big-data/
2020 project “Bridging the gap between artificial intelligence (accessed 29 September 2022).
Micheli et al. 5

European Commission (2018) Communication from the Commission: Minghini M, Sarretta A and Napolitano M (2022) Openstreetmap
Artificial Intelligence for Europe. 237. Luxemburg: European contribution to local data ecosystems in COVID-19 times:
Commission Publication Office. Experiences and reflections from the Italian case. Data 7(4): 39.
Floridi L, Cowls J, Beltrametti M, et al. (2018) AI4People—An Morley J, Floridi L, Kinsey L, et al. (2020) From what to how: An
Ethical framework for a good AI society: Opportunities, initial review of publicly available AI ethics tools, methods and
risks, principles, and recommendations. Minds and Machines research to translate principles into practices. Science and
28(4): 689–707. Engineering Ethics 26(4): 2141–2168.
Gevaert CM, Carman M, Rosman B, et al. (2021) Fairness and Mulder F, Ferguson J, Groenewegen P, et al. (2016) Questioning
accountability of AI in disaster risk management: Opportunities big data: Crowdsourcing crisis data towards an inclusive
and challenges. Patterns 2(11): 100363. humanitarian response. Big Data & Society 3(2): 1–13.
Gevaert CM, Sliuzas R, Persello C, et al. (2018) Evaluating the soci- Nativi S, Mazzetti P and Craglia M (2021) Digital ecosystems for
etal impact of using drones to support urban upgrading projects. developing digital twins of the earth: The destination earth
ISPRS International Journal of Geo-Information 7(3): 91. case. Remote Sensing 13(11): 2119.
Goodchild MF (2007) Citizens as voluntary sensors: Spatial data NSW Government (2020) Artificial Intelligence (AI) ethics policy:
infrastructure in the world of web 2.0. International Journal Mandatory ethical principles for the use of AI. Available at:
of Spatial Data Infrastructures Research 2: 24–32. https://www.digital.nsw.gov.au/policy/artificial-intelligence/
Goodchild MF (2021) Geoethics in the geospatial community. artificial-intelligence-ai-ethics-policy/mandatory-ethical (accessed
ArcNews 43: 30–31. https://www.esri.com/about/newsroom/ 25 April 2022).
arcnews/geoethics-in-the-geospatial-community Owusu M, Kuffer M, Belgiu M, et al. (2021) Towards user-driven
Goodchild MF, Guo H, Annoni A, et al. (2012) Next-generation earth observation-based slum mapping. Computers, Environment
digital earth. Proceedings of the National Academy of and Urban Systems 89: 101681.
Sciences 109(28): 11088–11094. Poom A, Järv O, Zook M, et al. (2020) COVID-19 is spatial:
Gore A (1998) The Digital Earth: Understanding our planet in the Ensuring that mobile big data is used for social good. Big
21st Century. Speech written for delivery at the opening of the Data & Society 7(2): 205395172095208.
California Science Center, Los Angeles. Sadowski J, Viljoen S and Whittaker M (2021) Everyone should
Haklay M (2016) Why is participation inequality important? In: decide how their digital data are used — not just tech compan-
Capineri C, Haklay M, Huang H, et al. (eds) European ies. Nature 595(7866): 169–171.
Handbook of Crowdsourced Geographic Information. London, Shanley LA, Tano M, Gabrynowicz JI, et al. (2015) Spatial Data
UK: Ubiquity Press, pp. 35–44. Sovereignty and Privacy in Indian Country. University of
Haugh HM (2021) The governance of entrepreneurial community Wisconsin-Madison. Available at: https://zenodo.org/record/
ventures: How do conflicting community interests influence 4085173#.X4XBaC2ZPxo (accessed 29 September 2022).
opportunity exploitation? Journal of Business Venturing Taylor L (2017) What is data justice? The case for connecting
Insights 16: e00265. digital rights and freedoms globally. Big Data & Society
Ibrahim RE, Elramly A and Hassan HM (2021) Open systems 4(2): 1–14.
science: Digital transformation and developing business UNESCO (2021) Recommendation on the Ethics of Artificial
model toward smart farms’ platform. International Journal of Intelligence. France: UNESCO. Available at: https://unesdoc.
Circuits, Systems and Signal Processing 14: 1054–1073. unesco.org/ark:/48223/pf0000380455 (accessed 3 November
ISDE (2012) International Society for Digital Earth website. 2022).
Available at: http://www.digitalearth-isde.org/list-90-1.html Vesnic-Alujevic L, Nascimento S and Pólvora A (2020) Societal
(accessed 29 September 2022). and ethical impacts of artificial intelligence: Critical notes on
Jobin A, Ienca M and Vayena E (2019) The global landscape of AI European policy frameworks. Telecommunications Policy
ethics guidelines. Nature Machine Intelligence 1(9): 389–399. 44(6): 101961.
Kyem Kwaku PA (2004) Of intractable conflicts and participatory Vespe M, Iacus SM, Santamaria C, et al. (2021) On the use of data
GIS applications: The search for consensus amidst competing from multiple mobile network operators in Europe to fight
claims and institutional demands. Annals of the Association COVID-19. Data & Policy 3(e8): 1–10. doi: 10.1017/dap.
of American Geographers 94(1): 37–57. 2021.9
Micheli M, Ponti M, Craglia M, et al. (2020) Emerging models of Wu W, Huang T and Gong K (2020) Ethical principles and govern-
data governance in the age of datafication. Big Data & Society ance technology development of AI in China. Engineering 6(3):
7(2): 205395172094808. 302–309.

You might also like