IGF 2019 OF #27 The Future of Artificial Intelligence and Sustainable Development

Description

Artificial intelligence (AI) offers a range of new opportunities to achieve the global Sustainable Development Goals. However, without the necessary capacity, technology and data to train AI, countries may miss out on the immense potential AI has to offer in the future. In addition, existing policy frameworks in many countries need to be updated to reflect the technological development of AI technologies. Against this background, the session launches the international project “FAIR Forward – Artificial Intelligence for All” by the German Ministry for Economic Cooperation and Development.

The project strives for a more open, inclusive and sustainable approach to AI on an international level. On this occasion, stakeholders from government, international organizations and the private sector present their work and projects on artificial intelligence including in the area of responsible AI, AI-driven voice interaction, mobile data and ethical policy frameworks.

Organizers

GIZ - Deutsche Gesellschaft für Internationale Zusammenarbeit
Robert Kirkpatrick, Director, UN Global Pulse Johannes Wander, Advisor Sector Programme Digital Development, GIZ Deutsche Gesellschaft für Internationale Zusammenarbeit Andreas Foerster, Head of Digital technologies for development unit, German Federal Ministry for Economic Cooperation and Development

Speakers

Audace Niyonkuru, CEO, Digital Umuganda

Michael Krake, Director, Private Sector, Trade, Employment, Digital Technologies, German Federal Ministry for Economic Cooperation and Development

Mila Romanoff, Lead Data Policy and Governance, UN Global Pulse 

Balthas Seibold, Head of FAIR Forward Sector Programme Digital Development, GIZ Deutsche Gesellschaft für Internationale Zusammenarbeit

Lukas Borkowski, European Partnerships Lead & Country Director Madagascar, Viamo

 

On-site moderator

Wadzi Motsi-Khatai

 

Online Moderator

Johannes Wander, Finn Hagemann

SDGs

GOAL 5: Gender Equality
GOAL 8: Decent Work and Economic Growth
GOAL 9: Industry, Innovation and Infrastructure
GOAL 10: Reduced Inequalities

1. Key Policy Questions and Expectations

Guiding policy questions for our speakers include:

  1. How can we foster international cooperation and peer-learning on the use of ethical artificial intelligence?
  2. How can we promote technical expertise and open training datasets for local artificial intelligence development?
  3. How can policy frameworks and governance approaches ensure the responsible development and use of artificial intelligence?
2. Summary of Issues Discussed

This Open Forum was jointly hosted by the German Federal Ministry for Economic Cooperation and Development and UN Global Pulse. BMZ unveiled its new project “FAIR Forward. It strives for a more open, inclusive and sustainable approach to AI. A speaker from Ruanda showed, how better open access to African language data can enable the development of AI-based voice interaction in local languages. The aim is to empower marginalized groups and enable local innovation. UNGP presented its work in Africa: Supporting political frameworks for a value-based AI and better data protection e.g. in Ghana and Uganda. There was broad support for the initiative, in particular on creating open datasets of marginalized languages. It was added that attention has to be paid to the choice of languages, particularly the amount of people who speak the language as well as the state of digital development of communities and accountability of AI systems.

3. Policy Recommendations or Suggestions for the Way Forward
  • Support initiatives to build open voice datasets of languages that the private sector does not cover or provide publicly for economic reasons to allow for inclusive AI development
  • Further focus on countries and regions that have been left behind in the development of AI technologies so far
  • Focus on languages that are spoken by many people i.e. Swaheli
  • Avoid exploitation of datasets and allow for equal and open access to the datasets
  • In terms of policy frameworks, new regulation has to be tailored for the local context and shall not be simply copied from other countries without adaptation.
4. Other Initiatives Addressing the Session Issues

Initiatives that support the strengthening of local knowledge and educational offers on AI development, creating freely available data for local AI innovation and helping to shape the political framework for a value-based AI and better data protection are:

  • BMZ FAIR Forward
  • UN Global Pulse
  • Digital Umuganda
  • VIAMO
  • Mozilla’s Common-Voice project
  • IDRC, Knowledge for All
  • SIDA: Artificial Intelligence for Development
  • UNESCO: Artificial intelligence with human values for sustainable development
5. Making Progress for Tackled Issues

In order to allow for an inclusive future of AI, the project “FAIR Forward – Artificial Intelligence for All” has been started and will run for three years. Mozilla’s Common-Voice project is underway to not only collect but also to make open-source voice databases on various languages publicly available. The project is continuing to seek active input and advice from the IGF ecosystem.

6. Estimated Participation

Total onsite: 110

Total online: 8

 

Please estimate the total number of women present onsite and online.

Total onsite: 32

Total online: Cannot indicate online number of women

7. Reflection to Gender Issues

Voice powered Artificial Intelligence allows to build services that better serve underprivileged groups and women. One example discussed at the session was VIAMOs 3-2-1 system of information on breastfeeding and child healthcare for mothers.