Q1/2019 - UNESCO

Paris, 4. März 2019

Konferenz „Principles for AI: Towards a Humanistic Approach?“

Auf einer hochrangigen Multistakeholder-Konferenz hat sich die UNESCO am 4. März 2019 mit dem Thema Künstliche Intelligenz auseinandergesetzt. An der Konferenz nahmen über 300 Experten aus 100 Ländern teil. Im Zentrum der Konferenz stand die Frage, wie die von der UNESCO verfolgten ROAM-Prinzipien für Internet Universalität (ROAM steht für Rights, Openess, Access und Multistakeholderism) für die Entwicklung von Politiken zur künstlichen Intelligenz angewandt werden können.

UNESCO-Generaldirektorin Audrey Azoulay reklamierte in ihrer Eröffnungsansprache eine führende Rolle der UNESCO bei der weiteren politischen Diskussion zu KI. Das „S“ im Namen der UNESCO stehe für „Science“. Die UNESCO als Wissenschaftsorganisation hätte daher eine ursächliche, aus ihrem Verfassungsauftrag resultierende Pflicht, sich mit diesem Thema zu beschäftigen[1]. Für die UNESCO sei es insbesondere wichtig, das die Entwicklung und Anwendung künstlicher Intelligenz nicht zur Aushöhlung, sondern zur Stärkung der Menschenrechte beiträgt, insbesondere den Rechten auf Meinungsäußerungsfreiheit, Schutz der Privatsphäre und gleichberechtigte Teilhabe am öffentlichen Leben.

José Ángel Gurría,Generalsekretär der OECD (Organisation für wirtschaftliche Zusammenarbeit und Entwicklung),warb für eine enge Vernetzung der verschiedenen internationalen KI-Initiativen. Er bot der UNESCO eine enge Kooperation mit der OECD an und schlug vor, dies mit den kanadischen und französischen Aktivitäten im Rahmen der G7 sowie weiteren Expertengruppen, z.B. in der IEEE, zu verbinden. Er sprach sich dafür aus, dass IGF in Berlin 2019 als eine Plattform zur Fortsetzung der Diskussion zu nutzen[2].

In einem gesonderten Papier „Steering AI for Knowledge Societies“ hat die UNESCO detailliert aufgeschlüsselt, wie die ROAM-Prinzipien bei KI-Entwicklungen mit Leben erfüllt werden können. In einem Aktionsplan werden elf verschiedene Maßnahmen vorgeschlagen. Das Papier und die Ergebnisse der Konferenz wurden am 21. März 2019 in Genf während der 40. Tagung des UN-Menschenrechtsrates vorgestellt[3].

Mehr zum Thema
Q1/2019UN
  1. [1] Rede von UNESCO Generalsekretärin Audrey Azoulay auf der UNESCO-Konferenz "Principles for AI: Towards a Humanistic Approach?", Paris, 4. März 2019: „Comme les révolutions industrielles qui l’ont précédée, l’intelligence artificielle révolutionne notre rapport au temps et à l’espace. Mais cette révolution représente quelque chose de nouveau: non pas simplement parce qu’elle porte ces transformations à un degré sans précédent, mais parce qu’elle est sur le point de révolutionner, aussi, notre rapport à l’humain. Dès aujourd’hui, nous nous trouvons donc au moment où nous devons articuler d’une part la capacité de choix d’innovation technologique, sociale, économique, et d’autre part la responsabilité de choix éthiques.Il ne s’agit pas d’avoir peur, ni d’être naïf: il s’agit d’être pleinement conscient de notre responsabilité. Une responsabilité double : comprendre d’abord ce qui est à l’œuvre, puis définir le chemin de crête qui nous permettra de mettre l'intelligence artificielle au service du bien commun, tel que nous l’avons collectivement défini dans l’Agenda 2030 des Nations unies, et au-delà. ….,Although estimates vary, we know that certain categories of jobs will disappear and others will be created, but there is a huge risk that many will be left behind.More than ever, lifelong learning will be crucial. Through education we can transform the world. But in order to play that role, to allow young generations of today and tomorrow to act in the world and on the world, education must first transform itself.In this regard, artificial intelligence opens up unprecedented prospects. ….This means that to the question of what kind of artificial intelligence we want to develop, we must add another one: how to prepare humanity to live with artificial intelligence? What tools, what education do we need to help future generations grow up in a world transformed by artificial intelligence? Should computer coding be taught from an early age, as a universal grammar? How could we make a better use of the fundamental role of humanities? …Car ne pas s’occuper de ces choix, c’est laisser libre cours à la concurrence la plus dure sans règle du jeu. Il est certainement prématuré de vouloir réglementer ce sujet au niveau mondial, mais il est plus que temps de définir un socle de principes éthiques qui encadreraient cette disruption technologique. Pour qu’elle soit au service des préférences collectives. Pour qu’elle repose sur des valeurs humanistes. La confiance dans ces innovations se gagnera au prix de cet effort collectif. Transparence, loyauté, élimination des biais, notamment de genre – car le manque de femmes dans les métiers de l’IA accroit les inégalités – responsabilité, inclusion, débat sur les règles imbriquées dans les algorithmes : autant de questions qui doivent être traitées ab initio et non pas après que les usages et les développements seront déjà là.“ Siehe: https://unesdoc.unesco.org/ark:/48223/pf0000366972?posInSet=3&queryId=N-995fe4da-8d83-4b87-a7ca-00b62ebf42c4
  2. [2] José Ángel Gurría, OECD-Generalsekretär, Rede auf der UNESCO-Konferenz zur künstlichen Intelligenz, Paris, 4. März 2019: „Today’s conference looks at a key aspect of that digital transformation: Artificial Intelligence. It examines how to bring human-centred values, inclusion and well-being to the heart of AI. In other words, how to make AI less artificial and more intelligent. AI is reshaping economies, facilitating better lives and helping people make better predictions and decisions It is already present in our everyday lives. The smartphone in your pocket can use AI to detect possible health issues, or speech recognition to offer on-the-spot translation. Investment in AI reflects the optimism people have for this technology. AI start-ups attracted over 12% of all worldwide private equity investments in the first half of 2018. At the same time, AI technologies are still in their infancy. Much potential remains to be fulfilled, particularly in the service of people and the planet. AI also brings challenges and uncertainties All the more reason to think carefully about how to ensure responsible stewardship of AI. We have to get this right, because while AI is driving optimism, it is also fuelling anxieties and ethical concerns. In particular, there are questions around the trustworthiness of AI systems, including the dangers of codifying and reinforcing existing biases – such as those related to gender and race – or of infringing on human rights and values such as privacy. We must ensure we avoid establishing automatised discrimination in hiring processes, for example, or in the criminal justice system. We need to design AI systems that are transparent and accountable. The OECD is contributing to define the rules of the game No single country or category of actors has all the answers to these challenges. We need a global multi-stakeholder response to a global issue. In 2018 the OECD formed an expert group to scope a set of principles to facilitate innovation, adoption of and trust in, AI. Just a few weeks ago in Dubai, that group reached agreement on the core values and actions required – an immense achievement, which is inspiring our work on an OECD Recommendation for adoption by governments. The Recommendation will represent a global reference point underpinning responsible stewardship of trustworthy AI. At its core is a focus on inclusive, sustainable growth and well-being, human-centred values and fairness, transparency, robustness and accountability. It will identify critical actions by governments, including investment in AI research and development, fostering of digital and innovation ecosystems, and an agile policy environment, building human capacity and preparing for job transformation. Importantly, the principles also call for international co-operation. We must progress together on AI-related technical, ethical and legal issues, to foster alignment of standards and codes of conduct, and inter-operability of laws and regulations. We are already bringing our measurement and analytical expertise to bear. This year we will establish an AI Policy “Observatory”, a participatory hub that will help provide national governments with the information they need to make evidence-based decisions. Our goal is to ensure consistency and complementarity between OECD work and other international initiatives, including under the G20 and G7, the European Commission, organisations such as the IEEE, and the UN, including importantly UNESCO itself. Our Observatory will seek to combine the resources of the OECD with partners from other international organisations and stakeholder groups. This could be a vehicle for the OECD and UNESCO to collectively support a Franco-Canadian initiative to form an “International Study Group on Artificial Intelligence”, leveraging our mutual expertise on ethical, technical and scientific issues on AI, as demanded by President Macron last year at the Internet Governance Forum.“ http://www.oecd.org/about/secretary-general/unesco-conference-on-principles-for-ai-towards-a-humanistic-approach-march-2019.htm
  3. [3] Siehe: Steering AI for Knowledge Societies: A ROAM Perspective; Chapter Overarching Options: „All stakeholders can consider addressing AI in relation to communication-information through: 1. Using UNESCO’s Internet Universality indicators to measure human Rights, Openness, Accessibility, and Multi-stakeholder participation and to thereby map and improve the ecosystem in which AI is developed, applied and governed. 2. Applying human rights norms that can inform more specific guidelines for rights to expression, privacy, and participation in public life. 3. Raising awareness of ownership and access to big data, AI skills and technologies, and the issues of who benefits, as well as harms such as marginalization or manipulation of human agency. 4. Assessing algorithmic discrimination in order to protect the right to equality of all, in particular of historically marginalized populations. 5. Participating in interdisciplinary research on how AI intersects with human rights, openness, accessibility and multistakeholder governance, and promoting Open Access publishing of the research results. 6. Upholding open market competition to prevent monopolization of AI, and requiring adequate safeguards against violation of ethical practices by market actors. 7. Facilitating development of norms and policies for improving openness and transparency in AI algorithms through elements of ex-ante information disclosure and ex-poste monitoring of algorithmic decision-making. 8. Working to reduce digital divides, including gender divides, in regard to AI access, and establishing independent monitoring mechanisms. 9. Motivating for more active participation in AI governance from all stakeholder groups, including but not limited to Governments, the Private Sector, Technical Community, Civil Society, Academia, International organizations and Media. 10. Ensuring gender equality, linguistic and regional diversity as well as the inclusion of youth and marginalized groups in multi-stakeholder dialogues on AI issues. 11. Integrating discussion of AI issues into relevant events such as international days around press freedom, disability, and universal access to information, and drawing in networks such as UNITWIN, Orbicom, Gapmil, and Gamag, as well Category 2 institutes, NGOs, IFAP national committees and UNESCO National Commissions.“ Siehe: https://en.unesco.org/news/unesco-unveils-summary-research-ais-implications-human-rights-40th-human-rights-council