IGF 2020 WS #125 How do you embed trust and confidence in AI?

Information

Session

Time:

Thursday, 12 November, 2020 - 12:20 to 13:20 UTC

Room:

Room 2

About this Session:

The proposed session will debate questions regarding governance methods to providing citizens, organizations and governments with the means to assess the trustworthiness of AI systems. As AI is playing an increasingly vital role in the services that are provided over the internet, as well as running and maintaining the core infrastructure that the internet depends on, trust in the Internet is increasingly defined by trust in these AI systems.

Area:

Data

Topic(s):

And Other Regulatory or Non Regulatory Models For Data Governance
Artificial Intelligence
Sustainable Business Models

Organizer 1: Ansgar Koene, EY
Organizer 2: Fatima Hassan, EY

Speaker 1: Clara Neppel, Technical Community, Western European and Others Group (WEOG)
Speaker 2: Parminder Jeet Singh, Civil Society, Asia-Pacific Group
Speaker 3: Ansgar Koene, Private Sector, Western European and Others Group (WEOG)
Speaker 4: Yohko HATADA, Civil Society, Asia-Pacific Group
Speaker 5: Amani Abou-Zeid , Intergovernmental Organization, African Group

Additional Speakers:

The moderator role will be take over by Catherine (Katie) Kummer, Global Deputy Vice Chair, Public Policy at EY, who will replace Hassan-Szlamka Fatima in this role since Fatima will be on maternity leave.

Mr. Ajijola Abdul-Hakeem,the Chair of the African Union Cybersecurity Expert Group (AUCSEG), will replace Dr. Amani Abou-Zeid, who unfortunately will have to attend to other AU matters at the time of our panel.

Moderator:

Ansgar Koene, Private Sector, Western European and Others Group (WEOG)

Online Moderator:

Fatima Hassan, Private Sector, Western European and Others Group (WEOG)

Rapporteur:

Ansgar Koene, Private Sector, Western European and Others Group (WEOG)

Format:

Debate - Auditorium - 60 Min

Policy Question(s):

Governance dimensions for data-driven technologies • What is the relationship between ethical considerations, Human Rights and trust in business use of emerging data driven technologies? • What societal and economic benefits are enabled by implementing business processes for monitoring and reporting AI system trustworthiness performance? How should these benefits be weighed against the need to protect competitive advantages/IPR? Data-driven emerging technologies • How do we move from the articulation of ethical and human rights principles for AI to the operationalization of those in business practices around the deploying of AI technologies? • How could data driven business practices benefit policy-making through data and analytical capability sharing agreements? Data-driven business models • How to respect privacy and agency over the use of data from individuals and businesses without sacrificing the beneficial potential of secondary uses of data for machine learning? What are the technological or regulatory strategies to address this? • How can data governance help to mitigate power imbalences between global and local economic actors? Data access, quality, interoperability, competition & innovation • How can we ensure equitable access to data and compute infrastructure for fostering competition and innovation? • How can we ensure portability and interoperability of data for fostering innovation with a Human Rights approach? • How is data quality playing a role in the conception of trust in the use of AI systems?

Issues, Challenges and Opportunities addressed:

This session will seek to answer the following questions: - What are the key attributes of a trusted AI systems and outcomes? - What are the best practices businesses should consider when designing and deploying AI tools in relation to data? (e.g. oversight of how data is used, trust in the data being used and data security) - What employee training or skillsets are needed to continuously monitor and integrate the relevant technical and ethical attributes into AI enterprise platforms? - How should businesses engage with the community of stakeholders affected by the use of AI systems? - What role should Standards, Certification and Audits play in establishing trustworthiness of AI systems?

SDGs:

GOAL 4: Quality Education
GOAL 5: Gender Equality
GOAL 8: Decent Work and Economic Growth
GOAL 9: Industry, Innovation and Infrastructure
GOAL 10: Reduced Inequalities
GOAL 12: Responsible Production and Consumption

Description:

Artificial intelligence (AI) technologies are rapidly outpacing the organizational governance and controls that guide their use. At the same time, organizations need to build trust with their internal and external stakeholders that AI systems are functioning reliably and accurately, and they need to be able to trust the data being used. Amid these considerations, it is increasingly clear that failure to adopt globally consistent governance and ethical standards that foster trust in AI will limit organizations’ ability to harness the full potential of these exciting technologies to fuel future growth.

Expected Outcomes:

1) Facilitate the debate as well as shaping the evolution of norms, principles, best practices of business use of AI system and digital data. 2) Identify differing viewpoints regarding governance approaches to the use of AI systems. 3) Policy recommendations and key messages report to the IGF community. 4) Foster greater collaboration amongst stakeholder from public, private and civil-society sectors regarding the governance of AI systems and how it intersects with data protection.

Relevance to Internet Governance: The proposed session will debate questions regarding governance methods to providing citizens, organizations and goverments with the means to assess the trustworthiness of AI systems. As AI is playing an increasingly vital role in the services that are provided over the internet, as well as running and maintaining the core infrastructure that the inetner depends on, trust in the Internet is increasingly defined by trust in these AI systems.

Relevance to Theme: As highlighted by the policy questions that this session addresses, the session contributes to the thematic track on Data on four of the six subtracks: 1) Governance dimensions for data-driven technologies 3) Data-driven emerging technologies 4) Data-driven business models 5) Data access, quality, interoperability, competition & innovation

Discussion Facilitation:

The session will be opened by the onsite moderator to provide participants an overview of the policy questions discussed in the session, the professional background of the speakers, and the format of interaction. The moderator will ensure the audience from both offline and online will be able to ask questions to the speakers immediately following their opening statement to encourage active participation. In the second part, the session will move to the discussions and debate. The moderator will invite each speaker to express their views on a set of questions and guide the debate amongst speakers and the audience to foreground their common ground and differences. The workshop organizers and moderators will discuss the content of questions with speakers in advance to ensure the quality and flow of the discussion and debate. In the third part, moderators will invite questions from the audience and online participants, the question time will last about 30 minutes in order to provide sufficient interactions amongst speakers, audience and online participants. Online participants will be given priority to speak, and their participation will be encouraged by moderators. The onsite moderator will summarise the findings and recommendations and future actions of the panel.

Online Participation:

Usage of IGF Official Tool.

Main menu

IGF 2020 WS #125 How do you embed trust and confidence in AI?

Information