Trustworthy AI in practice: an analysis of practitioners' needs and challenges
Recently, there has been growing attention on behalf of both 9 academic and practice communities towards the ability of 10 Artificial Intelligence (AI) systems to operate responsibly and 11 ethically. As a result, a plethora of frameworks and guide- 12 lines have appeared to support practitioners in implementing 13 Trustworthy AI applications (TAI). However, little research 14 has been done to investigate whether such frameworks are 15 being used and how. In this work, we study the vision AI 16 practitioners have on TAI principles, how they address them, 17 and what they would like to have – in terms of tools, knowl- 18 edge, or guidelines – when they attempt to incorporate such 19 principles into the systems they develop. Through a survey 20 and semi-structured interviews, we systematically investi- 21 gated practitioners’ challenges and needs in developing TAI 22 systems. Based on these practical findings, we highlight rec- 23 ommendations to help AI practitioners develop Trustworthy 24 AI applications.
Wed 19 JunDisplayed time zone: Amsterdam, Berlin, Bern, Rome, Stockholm, Vienna change
11:00 - 12:20 | Human AspectsShort Papers, Vision and Emerging Results / Research Papers / Industry at Room Capri Chair(s): Guilherme Horta Travassos Federal University of Rio de Janeiro | ||
11:00 15mTalk | Trustworthy AI in practice: an analysis of practitioners' needs and challenges Research Papers Maria Teresa Baldassarre Department of Computer Science, University of Bari , Domenico Gigante SER&Practices and University of Bari, Azzurra Ragone University of Bari, Sara Tibidò Scuola IMT Alti Studi Lucca, Marcos Kalinowski Pontifical Catholic University of Rio de Janeiro (PUC-Rio) | ||
11:15 15mTalk | "Looks Good To Me ;-)": Assessing Sentiment Analysis Tools for Pull Request Discussions Research Papers Daniel Coutinho Pontifical Catholic University of Rio de Janeiro (PUC-Rio), Luísa Cito Pontifical Catholic University of Rio de Janeiro, Maria Vitória Lima Pontifical Catholic University of Rio de Janeiro (PUC-Rio), Beatriz Arantes Pontifical Catholic University of Rio de Janeiro (PUC-Rio), Juliana Alves Pereira PUC-Rio, Johny Arriel PUC-Rio, João Godinho Pontifical Catholic University of Rio de Janeiro (PUC-Rio), Vinícius Martins PUC-Rio, Paulo Vítor C. F. Libório Pontifical Catholic University of Rio de Janeiro (PUC-Rio), Leonardo Leite Federal University of Alagoas (UFAL), Alessandro Garcia Pontifical Catholic University of Rio de Janeiro (PUC-Rio), Wesley Assunção North Carolina State University, Igor Steinmacher Northern Arizona University, Augusto Baffa PUC-Rio, Baldoino Fonseca Federal University of Alagoas | ||
11:30 15mTalk | Motivation Research Using Labeling Functions Research Papers DOI Pre-print | ||
11:45 15mTalk | Insight AI Risk Detection Model - Vulnerable People Emotional Situation Support Industry Diego Gosmar Open Voice Trustmark Ethical use task force Linux Foundation AI & DATA, Elena Peretto Fundaci— Ajuda i Esperanچa, Oita Coleman Open Voice Trustmark Ethical use task force Linux Foundation AI & DATA | ||
12:00 10mTalk | On the Use of ChatGPT for Code Review Short Papers, Vision and Emerging Results Miku Watanabe Nara College, National Institute of Technology/Nara Institute of Science and Technology, Yutaro Kashiwa Nara Institute of Science and Technology, Bin Lin Radboud University, Toshiki Hirao , Ken'Ichi Yamaguchi , Hajimu Iida Nara Institute of Science and Technology Pre-print | ||
12:10 10mTalk | What You Use is What You Get: Unforced Errors in Studying Cultural Aspects in Agile Software Development Short Papers, Vision and Emerging Results Michael Neumann University of Applied Sciences & Arts Hannover, Klaus Schmid University of Hildesheim, Lars Baumann DOI Pre-print | ||