IGF 2022 WS #354 Affective Computing: The Governance challenges

Time
Tuesday, 29th November, 2022 (12:05 UTC) - Tuesday, 29th November, 2022 (13:05 UTC)
Room
CR5

Organizer 1: Diogo Cortiz da Silva, Network Information Center (NIC.br)
Organizer 2: Lucia Santaella, Pontifical Catholic University of São Paulo
Organizer 3: Hartmut Richard Glaser, Brazilian Internet Steering Committee - CGI.br

Speaker 1: Diogo Cortiz da Silva, Technical Community, Latin American and Caribbean Group (GRULAC)
Speaker 2: Lisa Feldman Barrett, Civil Society, Western European and Others Group (WEOG)
Speaker 3: Javier Hernandez, Private Sector, Western European and Others Group (WEOG)
Speaker 4: Jessica Szczuka, Civil Society, Western European and Others Group (WEOG)
Speaker 5: Marina Meira, Civil Society, Latin American and Caribbean Group (GRULAC)

Moderator

Henrique Xavier, Technical Community, Latin American and Caribbean Group (GRULAC)

Online Moderator

Gabriela Nardy, Technical Community, Latin American and Caribbean Group (GRULAC)

Rapporteur

Pollyanna Rigon, Technical Community, Latin American and Caribbean Group (GRULAC)

Format

Round Table - U-shape - 60 Min

Policy Question(s)

This workshop aims to establish an initial and interdisciplinary discussion to introduce the topic of Affective Computing on the IGF agenda, so we propose the following policy questions: To what extent Affective Computing is a reliable and trustworthy technology to infer about a user's emotions? What are the scientific and technical limits for it and how could we use it to improve quality of life as well as monitor global health risks on a large scale? Affective Computing solutions are typically designed to be global, how can we ensure that cultural and local criteria are taken into account in the design phase to enable equitable access to technology? How can we make sure that Affective Computing is not developed and used for harmful purposes? What values and norms should guide the development and use to enable this?

Connection with previous Messages:

SDGs

3. Good Health and Well-Being
10. Reduced Inequalities

Targets: This workshop proposal is related to the following SDGs: 3- Good Health and Wellbeing One topic that we will address in this workshop is the potential use of Affective Computing (if used properly) to improve quality of life as well as monitoring global health risks on a large scale. 10 - Reduce inequality within and among countries This SDG is linked with one of our policy questions regarding how we can ensure that cultural and local criteria are taken into account in the design phase to enable equitable access to affective computing technologies.

Description:

Affective Computing is an emerging technology that comprises the study of how computers can recognize, interpret and simulate human affections. The approach employs various types of input, such as images of facial expressions, text, voice, and physiological data. With the advancement of Artificial Intelligence (AI), Emotion Recognition is becoming more detailed and fine-grained. A few years ago the most common task was classifying emotion in terms of its valence (positive, negative, and neutral), but today some techniques allow recognizing a user's emotional state into up to 40 different categories. If we take into account the advances in Virtual Reality (VR) as an enabling technology for the Metaverse, we could find an even more favorable environment for collecting and mapping data about people's subjectivity, cognition, and emotions. This scenario can have different practical results, such as providing a better user experience or empowering people by promoting a richer understanding of their emotional states, providing a tool for monitoring global health risks on a large scale. However, they come with new privacy and governance challenges. This workshop will discuss the growing adoption of AI models to identify users' emotions that could indicate highly accurate subjectivity aspects, for multiple purposes, such as in profiling or to deliver a better and customized experience to the users on the Web and in the Metaverse in the future. Despite the convincing results of those systems, in scientific terms, there is still no consensus on the technical, social, and ethical feasibility of using AI to infer peoples' emotions. This workshop will provide an interdisciplinary space to bring together experts from neuroscience, AI, privacy, and regulatory matters to discuss and identify the potential use of those technologies and possible risks to drive future governance needs.

Expected Outcomes

Inform audiences and all interested parties about key uses, new trends, and challenges of Affective Computing. Provide a list of emerging specific initiatives to ensure ethical and safe use of affective computing on the web and the Metaverse Elaborate key recommendations from speakers and participants to introduce a multi-stakeholder perspective of Affective Computing to guide the local policy agenda.

Hybrid Format: The session will be divided into three main segments: introduction, debate and interaction. The first segment will consist of the initial presentation of each of the participants (duration of 5 minutes for each participant) to bring a multistakeholder view to the topic. The second segment will consist of the speakers' view on the policy questions. The last segment will consist of the interaction between the audience and the speakers. For this final part, the Q&A questions will be collected by the online moderator and given to the in-person moderator, who will select and distribute them among the speakers. We plan an online meeting with all the speakers one-week before the IGF to coordinate the interventions of the speakers based on who will participate online or in person. The onsite moderator will then be able to coordinate the speaking time of each participant according to the topic and type of presence (online or onsite) to ensure the best possible experience for both onsite and online audience. We plan to use all tools provided by the IGF organization.

Online Participation

 

Usage of IGF Official Tool.