Navigating Ethical Horizons: Insights from TechEthos Policy Event

4 December 2023

Authored by: Giuseppe Loveno Garofalo 
Reviewed by: Greta Alliaj and Clara Boissenin

The TechEthos Policy Event held in Brussels on 14 November 2023 brought together high-level experts, including EU policymakers, researchers from academia, and industry representatives to discuss the ethical governance of emerging technologies in the digital transformation and green transition. The event was moderated by science journalist and broadcaster Vivienne Parry and drew over 60 on-site and 24 virtual participants. With the mission to discuss the ethical implications of emerging technologies, it kicked off with the opening remarks of policy analyst Mihalis Kritikos (DG RTD) emphasizing the evolution of tech-related EU policies and the need of a just transition. Next, TechEthos coordinator, Eva Buchinger (AIT) illustrated the TechEthos vision and Ethics by Design approach.

Ethics for the Digital Transformation

The morning session delved into ethical considerations for Digital Transformation, with Laura Weidinger, Senior Research Scientist at DeepMind setting the scene for an insightful panel discussion. Highlighting AI Safety, especially Generative AI, during her keynote speech, Weidinger stressed:

• The need for a sociotechnical approach in AI development which recognises potential repercussions of algorithms on individuals. To reach sociotechnical safety, developers must assess AI capability, human interaction, and potential side-effects after full deployment;
• EU’s AI Act should proactively advocate for a balanced distribution of responsibility, emphasizing foresight and pre-market testing implementation.

The session extended to explore ethical, social, and regulatory challenges of Digital Extended Reality through a dynamic panel discussion featuring Laura Weidinger, Alexei Grinbaum (CEA), Kevin MacNish (SOPRA STERIA), Alina Kadlubsky (Open AR Cloud Europe) and Ivan Yamshchikov (CAIRO). Key topics included:

• Challenges in implementing Ethics by Design in XR applications. The creation of digital ‘twins’ via XR often produces static representations that may not capture the dynamic real-world evolution. Also, the integration of AI, visualization, and biometric data in human-machine interactions adds complexity, presenting ethical challenges in XR design and implementation;
• Potential certification for development companies conducting impact assessments for their XR models;
• Ethical concerns related to who shapes AI companions and their actions – raising debates over accountability;
• The challenge of conceptual disentanglement in interactions between humans and AI companions, especially those with human-like attributes. These evolving relationships are understudied but may become a crucial focus in the future;
• The absence of a one-size-fits-all rule complicates the implementation of the Ethics by Design approach.

Ethics for the Green Transition

In the afternoon session, Behnam Taebi, full Professor of Energy & Climate Ethics at Delft University of Technology addressed the urgent need for achieving net-zero by 2050, exploring the ethical controversies surrounding climate engineering techniques such as Solar Radiation Management (SRM) and Carbon Dioxide Removal (CDR). Key takeaways included:

• SRM offers substantial benefits in reducing the pace of temperature and sea level rise, but its impact on ozone depletion, drought, health, and agriculture remains unknown;
• Advocacy for a “risk vs. risk” approach, weighing the risks of implementing climate engineering against the risks of not taking any action;
• Recognition of the need for private funding to address climate challenges while emphasizing the importance of business models that transcend profit-centric approaches.

The keynote speech was followed by a panel discussion on the ethical, social, and regulatory challenges featuring Behnam Taebi, Dominic Lenzi (University of Twente), Dušan Chrenek (DG CLIMATE ACTION), and Matthias Honegger (Perspectives Climate Research). Key points discussed encompassed:

• The need for a global collaboration for SRM research which incorporates principles of global justice, human rights protection, and legitimacy;
• The necessity for effective and legitimate governance of research;
• Warning about potential overreliance on the CDR technology – which might hinder emissions reduction efforts;
• The need to clarify how CDR can be implemented in accordance with the EU principles, the biodiversity strategy 2030, and the UNFCCC’s principle of common but differentiated responsibilities and respective capabilities;
• The necessity to scrutinize the role of fossil fuel industry in CDR deployment.

Conclusion

The conference ended with some final words from Maura Hiney (UCD Institute for Discovery), who contextualised the TechEthos project within the ALLEA Code of Conduct – a high-level framework for achieving high-quality, trustworthy, and responsible research. Finally, Eva Buchinger, Laurence Brooks (University of Sheffield), and Renate Klar (EUREC) provided some context concerning the project – seen as a continuation of previous projects and a foundation for future research on the ethics of emerging technologies.

The TechEthos Policy Event attempted to explore the ethical implications of the digital transformation and green transition, unravelling challenges that define our technological future. In the convergence of innovation and moral responsibility, a clear call emerges for global discourse. It is in these considerations that we shape ethical frameworks, ensuring that technology serves humanity with integrity. The echoes of the discussions held should resonate, urging us to collectively navigate the ethical landscapes that lie ahead.

Do you want to find out more? Take a look at the videos of the event on our YouTube Channel!

Share:

Highlighting key ethical issues determined from scenario creation, expert engagement and citizen engagement via game-based methodology and workshops

12 September 2023

Authored by: Robyn Sahota (Ecsite)
Reviewed by: Greta Alliaj (Ecsite) and Wenzel Mehnert (AIT)

Introduction to the multi-stage and multi-stakeholder scenario methodology

This article, summarizing the report D3.1, discusses the involvement of various stakeholders, including the research community and the general public, in considering the ethical implications of new and emerging technologies. This process is outlined as a multi-stage and multi-stakeholder approach within the TechEthos project. The approach involves three stages: creating basic technology scenarios, enriching these scenarios through engagement with experts, and further enhancing them through engagement with the public. This method aims to facilitate the expression of stakeholders’ attitudes, values, and expectations regarding the ethical dimensions of uncertain technologies.

The methodology for societal acceptance analysis includes mapping innovation ecosystems for selected technology families based on horizon scanning, creating basic scenarios for each technology family, and developing scenario exercises and games (see: the ‘TechEthos game: Ages of Technology Impact’). This is done in collaboration with science engagement professionals, gathering expert opinions on ethical implications through the basic scenarios, and obtaining public perspectives on ethical implications through engagement with citizens from various European countries. This approach ensures a comprehensive understanding of different stakeholder viewpoints.

The mapping of innovation ecosystems involves identifying stakeholders associated with three chosen technology families: climate engineering, digital extended reality, and neurotechnologies. The report provides brief descriptions of each family, its relevant technologies and case studies, and the stakeholders involved. The mapping process considers the technological context, relevant stakeholders such as researchers, technology experts, legal and policy experts, ethics experts, civil society organizations, media, and EU-funded projects.

The development of basic scenarios using the STEEPV framework

To evaluate the results, the STEEPV heuristic was applied, to form a common ground between the three stages mentioned above (basic scenarios, expert engagement, citizen engagement). The STEEPV framework is an analytical tool designed to comprehend external environmental changes amid uncertainty, information overload, and disruptions. The acronym covers six dimensions: Social, Technological, Economical, Ecological, Political & Legal, and Human Values developments (see D3.1, section 6.1). These dimensions explore factors like demographics, innovation, market dynamics, ecological impacts, regulations, and cultural influences. By dissecting these facets, STEEPV offers a comprehensive view of the evolving external environment, aiding in understanding past, present, and potential future trends.

The process involves several steps. First, trends and driving forces related to three technology families are identified, focusing on short, medium, and long-term impacts. This is informed by literature review and the results of the ecosystem mapping, yielding an overview of future developments based on STEEPV factors. Second, key factors characterized by “high impact” and “high uncertainty” are chosen from the trend list. This differentiation aids in forming projections for potential futures. Third, projections are developed using a 2×2 matrix, placing each key factor in the center and selecting two dimensions of uncertainty. About 4 projections per key factor per technology result. Fourth, a contingency analysis clusters these projections across STEEPV factors to form raw scenarios, capturing diverse yet plausible potential futures. This yields 3 contrasting raw scenarios per technology. Fifth, partner validation and feedback refine the scenarios, resulting in 3 contrasting basic scenarios per technology.

The basic scenarios were then advanced through a consultation with experts, where they contemplated the ethical concerns linked with the technology families as depicted in the scenarios. The suggested expert solutions to these ethical matters are elaborated upon in document D3.5. Next, the citizens’ awareness and perspectives were considered in the format of science cafes and game workshops. In the remainder of this article we will discuss the general outcomes from this process.

Key takeaways per technology family:

Climate Engineering (CE)

In the social dimension, discussions revolved around inequality, equity, and global consequences, with experts focusing on geopolitical and environmental instability, while citizens expressed concerns about human health impacts. Safety and reliability were highly valued by citizens but received less attention in basic scenarios and expert opinions. In the technological dimension, effectiveness and efficiency were central themes, and both groups emphasized “techsolutionism,” prompting discussions on alternative solutions and the need for trustworthiness and transparency. The ecological dimension emphasized ecosystem health and biodiversity, with citizens favouring natural approaches. Economic discussions centered on growth and local impacts for citizens, while experts addressed power dynamics. Both groups emphasized distributive justice in the political dimension, and citizens also discussed liability concerns. In the value dimension, lifestyle changes and moving away from consumerism were suggested alternatives, with varying perspectives on “techsolutionism” between experts and citizens.

Digital Extended Reality (XR)

In the social dimension, fears of a social divide, increased mental health challenges, and lack of authentic human connections due to virtualization were prominent. Technologically, ensuring safety, reliability, and trustworthiness of the were crucial, while ecological concerns involved increased CO2 emissions, rare earth mining, and changes in urban forms due to virtualization. Economically, XR’s impact on education, labour markets, and work-life balance raised both excitement and concerns. The political dimension highlighted data protection, democracy issues, and the lack of regulation for digital companies. In the value dimension, concerns included dignity of posthumous data, enhanced social isolation, biases (e.g., gender, race) in technology, and the potential loss of authentic human connections and experiences.

Neurotechnologies (NT)

In the social dimension, human health is prioritized, but concerns arise about altering definitions of disabilities or enhancements, potentially leading to discrimination. Technology gaps between users and non-users in different fields of application, like labour markets and individual enhancement, must be addressed for equal accessibility. Autonomy and agency are crucial, focusing on avoiding manipulative influences and ensuring data privacy. The technological dimension stresses responsible use, safety, reliability, and data privacy. The ecological dimension emphasizes ecosystem health, energy consumption, maintenance, and sustainability. The economic dimension discusses private company power, and data use transparency. The political dimension raises concerns about responsible use, neurodiscrimination, governance of big data, and infringement of autonomy. The value dimension calls for a human-centered approach, considering diverse perspectives, including neurodiversity.

Conclusions

Three overarching themes emerged from the analysis: equity, reliability, and environmental sustainability.

1. Equity: The fair and just distribution of benefits and risks associated with emerging technologies. This includes considerations of accessibility, affordability, and fair distribution of benefits to prevent potential social divides and ensure responsible innovation.
2. Reliability: Concerns about the feasibility, safety, and accountability of technologies and their applications. It involves ensuring that technologies work as intended, with minimal unintended side-effects, and companies being held accountable for their actions if necessary.
3. Environmental Sustainability: Focusing on the minimal negative impact of technologies on the environment while promoting biodiversity and healthy ecosystems. This includes addressing climate change, pollution, resource depletion, and habitat destruction in the development and use of technologies, as well as rising energy consumption and CO2 emissions by providing a globally accessible infrastructure for digital applications.

The report’s multi-stakeholder approach showcased the complexity of ethical issues and the importance of considering diverse perspectives when responsibly developing emerging technologies. The findings contribute to the ongoing discussion on ethical and social impacts of these technologies and inform the operationalization of guidelines in future projects.

Read the full report

Share:

Citizen awareness and attitudes towards emerging technologies: key takeaways from engagement workshops

12 September 2023

Authored by: Robyn Sahota (Ecsite)
Reviewed by: Greta Alliaj (Ecsite) and Wenzel Mehnert (AIT)

An important perspective the TechEthos project wanted to highlight alongside expert opinions was the citizen perspective. To encourage participation and facilitate conversation, an interactive game (D3.2) was developed to discuss the ethical issues related to climate engineering (CE), digital extended reality (XR), neurotechnologies (NT), and natural language processing (NLP). The goal of this exercise was to understand citizens awareness and attitudes towards these emerging technologies to provide insight into what the general public finds important.

First Science cafes were held, providing a casual forum for conversation in order to give the public general knowledge and build interest about the three technology families. These science cafes were conducted in: Vienna (Austria), Liberec (Czech Republic), Bucharest (Romania), Belgrade (Serbia), Granada (Spain), and Stockholm (Sweden). These sessions lasted around 90 minutes and consisted of short pitches from invited speaks, Q&A sessions, icebreakers, and open conversation.  

Following these science cafes was then the scenario game workshops (D3.1, section 5) held in the aforementioned countries, for a total of 20 events. These workshops engaged a wide audience from varied backgrounds, with an average of 16 participants per workshop with a duration of four hours. These workshops involved an introduction, warm-up, game play, reflection and a post-participation survey. Most participants were motivated to join these workshops due to their curiosity about technology (47.1%), followed by the desire to be involved in how technology develops (16.2%). 40 participants said they took part to the workshop because of their passion for the technology (12.5%). The gender representation in participation was fairly balanced when considering the combined country data per technology family workshop. 

Awareness

Overall, there was an even distribution of awareness (very aware, somewhat aware, not really aware) among the technology families. 70% of the participants had heard of the technology families discussed, with affective computing being the least known field (48.9% not really aware) and chatbots being the most known field (47.9% very aware), both within the NLP family. The awareness of chatbots can be linked to the prevalence in the general public’s daily life (e.g., Apple’s Siri, Amazon’s Alexa, ChatGPT). In discussions about virtual reality (89% are somewhat or very aware) and brain-computer interfaces (74% are somewhat or very aware), Science-Fiction served as a prominent frame of reference, shaping awareness, perception, and value of these technologies. Looking at the specific technologies in each technology family, we see that the most prominent technologies in CE are nature-based carbon dioxide removal (CDR), in NT brain-computer interfaces, and in NLP and XR it is chatbots and virtual reality.

The least known technologies within the technology families are similar distributed. Every third participant has not heard of solar radiation management (SRM) and Engineered CDR in CE, all three technologies in NT, and affective computing and digital twins in NLP and XR.

Citizens’ attitudes

Nature-based CDR emerged as the most favored technology, being perceived as environmentally friendly and balanced with nature. Conversely, SRM raised significant concerns among participants due to potential unforeseen disasters and pollution.

Neuroimaging generated the most enthusiasm, promising enhanced disease diagnosis and prevention. However, the brain-computer interface elicited worries about manipulation and data protection.

Among NLP technologies, chatbots and text analysis excited participants, offering simplicity and improved work efficiency. On the other hand, affective computing raised concerns about misuse and data privacy.

In the XR family, virtual reality excited participants, providing immersive experiences of distant or historical places. Meanwhile, the metaverse caused apprehension due to potential loss of human connection and data privacy issues.

Acceptance

The results indicate that participants generally accept the presented technologies, as the votes for excitement outweigh the votes for concerns. However, it’s important to note that participants often hold both excitement and concerns simultaneously, making their acceptance more complex.

In the post-survey conducted after game exercises and group deliberations, over 70% of participants expressed excitement or high excitement about possible future developments of the technologies. Simultaneously, approximately 50% of participants conveyed concerns or slight concerns regarding the technologies’ future developments.

Accepting emerging technology is complex as it involves shaping the yet-to-be-created intervention. Participants accept technology if certain conditions (X) are met or issues (Y) are prevented. Their mixed feelings of excitement and concerns reveal the need to understand shared values when discussing acceptance of emerging technologies. Reflecting on citizens’ values helps to comprehend their acceptance and to develop technologies more responsible.

Citizen values

From the workshop comments the citizen value categories (D3.1, Table 24) were extracted through qualitative coding, allowing for comparisons across all workshops.

Each technology family exhibits distinct prominent values. In the NT domain, human health, safety, and responsibility are emphasized due to its focus on the brain and nervous system. NLP and XR prioritize authentic human connection, experience, and responsible use, considering their aim to simulate human interactions.

CE highlights ecosystem health, followed by safety, reliability, effectiveness, efficiency, and justice, given its focus on manipulating natural systems. Safety and reliability are important across all three families. Responsible use and accountability are vital in NT, NLP, and XR. Ecosystem health is a shared concern across all families.

Conclusion

These workshops explored public awareness, attitudes, and citizen values towards emerging technologies. The goal was to elicit citizens’ excitements and concerns to inform ethical guidelines, which are then discussed in WP5, and to involve a diverse audience, including vulnerable groups, often underrepresented in expert processes. Overall, the “TechEthos game: Ages of Technology Impact” facilitated open and creative discussions.

Common value categories emerged across the technology families through these discussions:

1. Safety and reliability: Participants expressed worries about unknown effects, potential dangers, and health impacts in CE, XR, and NT.
2. Equity, diversity, and inclusion: Concerns centered on global distributive justice, access for all social groups, and respecting neurodiversity.
3. Responsible use and accountability: Participants stressed the importance of accountability for potential disasters, consequences, and data privacy in CE, XR, and NT.

Read the full report

Share: