Hundreds of Artificial Intelligence enthusiasts tuned in online to the “European AI Excellence and Trust in the World” and hundreds of interested people gathered on-site last week on March 15-16 to learn about European AI projects in the frame of Expo Dubai.
The High-Level Forum on the EU Vision for Trustworthy AI in the World (HLF) was the key event of the three widely attended events under the EU AI activities in the EXPO Dubai, which was organised by the European Commission through the InTouchAI.eu project.
The event organisers opened their live broadcasting to all AI enthusiasts in the world and those who care and are concerned about how AI is being and will be developed, regulated, and adopted following the EU values of human-centred and the development of AI with ecosystems of trust and excellence.
The High-Level Forum
Following the key theme of the EU vision for trustworthy AI in the world, the HLF gathered leaders from the EU Commission, EU member countries, EU and international organisations and associations and world-renowned experts, civil groups, academics, innovators, scientists and artists to discuss ethical, legal, technical, and societal consideration and implications of the EU vision in the development and regulation of AI.
In an HLF welcome remark, Mr Andrea Matteo Fontana, EU Ambassador to the UAE argued for human-centric, sustainable, and trustworthy AI within the European AI vision. As the venue host of HLF, Mr Paolo Glisenti, Commissioner General of Italy for Expo 2020 Dubai spoke about the importance of AI diplomacy, which he has experienced in the past five months at the Expo. He highlighted the importance of international cooperation between countries, organisations, institutions and others; global governance of AI, which must be inclusive, respective of human rights and shared values. He also mentioned the needed scientific collaborations among universities, research institutions and academics, which saw some starting points at the Expo. The public-private partnership was among the keyways to develop and harness human-centric AI solutions and applications.
The importance of trustworthy AI was also echoed by Ms Anna Ascani, Under Secretary of State, Ministry for Economic Development, Government of Italy, citing that Italy approved its AI national strategy last year with a lot of legal, human rights and ethical considerations taken into account. She shared the AI focus in the coming years in her country and how the EU AI would ensure the safe and truly positive applications of AI in all aspects of society.
In her opening statement, Ms Gabriela Ramos, Assistant Director-General for Social and Human Sciences, United Nations Educational, Scientific and Cultural Organization (UNESCO) said “Europe's approach to addressing the biases and discriminations in innovative technology is well-aligned with UNESCO’s AI Recommendation” and that it is time to start talking about a full implementation of AI ethics.
Mr Brando Benifei, Member of the European Parliament, Special Committee on Artificial Intelligence in the Digital Age (AIDA), Co-Rapporteur AI Act talked about how the EU can innovate and be competitive in AI without sacrificing people's rights. "We want to put our values at the core of AI legislation in a way that fosters technology development, it shouldn't be a contrast.” "We can protect citizens and at the same time incentivise creativity and innovation. At the EU Parliament, we work on building the world’s first legislation on AI. We also need to work with the world to create AI ownership for people, an AI that people want," he said at the HLF.
The keynote speech “AI Ethics by Design” by Prof Aimee van Wynsberghe, Alexander von Humboldt Professor of Applied Ethics of Artificial Intelligence, University of Bonn brought the audience to more in-depth thoughts and perspectives about AI ethics and the concepts of ethics related to AI’s trustworthiness and sustainability. Apart from the human-centred AI, which ensures the principle of human rights and European values, Prof Wynsberghe emphasized the two key pillars in sustainable AI: “AI for sustainability” and “sustainability of AI”. These two pillars are needed to think about at the very initial stages of creating and developing AI solutions by creators and innovators. The relevant considerations from the pillars would assure the true development of sustainable AI systems and solutions, which reinforced the trustworthiness of AI in application and implementation. There are lots of considerations that needed to start even before deciding to pursue the AI solutions including the need to look at the environment as a starting point and how to measure, access and regulate the AI ethics to ensure sustainable AI.
Prof Wynsberghe concluded that AI brings new opportunities, and they are parts of the solutions for sustainability, yet, as with any technology, AI also has unintended consequences. Thus, a trustworthy AI based on strong ethical drivers helps all of us to uncover and understand the risks facing the technology. Approaches toward ethical AI could be a step toward regulation that also pays attention to the sustainability of AI including uncovering hidden costs and the demographic who are suffering; and the need to include environmental costs in future AI regulations.
The keynote outlined some food for thought in the HLF Panel that followed, which opened the discussions of the European trustworthy AI at international perspectives including those at the global South and industries.
The panel was moderated by Dr Juha Heikkilä, Adviser for Artificial Intelligence, DG Connect, European Commission. He said that the European trustworthy AI approach, with which many of the audience was already familiar, would enhance the uptake in the technology. “For the benefits of the technology to materialize, we need to have uptake. And for the uptake of AI, we need trust in this technology so that it can be used safely and with confidence,” said Dr Heikkilä, reasoning why the European trustworthy AI approach is the way to go.
The panellists shared their perspectives based on a series of questions asked by the moderator, giving the importance of contexts and conditions in many parts of the world in the EU’s willingness to lead the trustworthy AI in the international sphere. The panellist also shared their learnings and practical considerations in the science of AI, policy, social impacts, and implications in terms of AI compliance for industries and the real benefits of AI in business, AI standardisation, measurement, evaluations, inclusion, education, and training needs of AI for the public and for building up AI talents who are aware of all the ethical considerations while developing technological robust AI systems that provide users with trust in the technology.
Expert Workshops on A Human-Centric Approach to Global Challenges
The HLF agendas were also discussed in detail at the expert workshop on “AI for Sustainability” and “AI for Health” that gathered a dozen of speakers from EU AI projects, academics, researchers, and leaders of businesses to dig deeper on the ethical aspects of all stages of AI development. Each workshop showcased three ongoing projects: AI4Cities, Cybele, and I-NERGY in the sustainability workshop while DIH-HeRo, EUCanImage and GENOMED4All in the health workshop. These are the projects that demonstrate all the ethical and societal considerations in all processes of building their products and services before the actual adoption.
The AI for Sustainability workshop features three EU projects in energy, smart and sustainable cities, and precision agriculture. The projects showcase how AI is being developed and applied in the areas to ensure sustainable growth while achieving both digital and green transitions of the EU.
A panel of renowned academics and industries experts also shared their learnings and the way forward in such AI-enabled innovation in sustainable growth. The trust of AI by all stakeholders in any industry and sector was highlighted in most of the discussion. Panellists debated on the importance of transparency, awareness, and early engagement of the stakeholders to ensure AI technology and its goals can be used safely while communicating about the real impacts and implications to the environment and human systems.
Health is one of the pioneering areas in adopting AI solutions for healthcare systems, treatments, diagnostics, precision medicine, pharmaceutical products and many more.
The workshop displayed three examples of European AI excellence in healthcare, including robots, AI in oncology, and algorithms for diagnosing, treating, and predicting haematological diseases. A panel discussion with international experts will bring perspectives on the practical implications of AI in healthcare, including how to ensure that these technologies are developed with adequate consideration for their social and ethical aspects.
Having the immediate health impacts and implications on people, AI for health needs quality of data, rights to own and share or not personal data, hospital data are key to any consideration of usage of data in algorithms and healthcare solutions. Experts cited the importance of aspects and find synergies from regulators, business drivers, developers, adopters, and others to ensure AI technology is being adopted knowing the technologies were developed with “ethics by design”.
The panellists also discussed health assessment technologies by design and how can we understand that clinicians now understand AI decision-making so that they can work with the technology to ensure safe and trustworthy diagnostics, prognosis, and others. Health is also the area that the EU values, including ensuring human rights and the general data protection regulation, should be held with utmost care at every single stage of development and usage of the technology.
Also demonstrating European Trustworthy AI in actions, AI Walk was an activity introducing EU’s projects using AI technology in use cases, including the artistic perspectives. The AI Walk is a collection of AI cases from different EU pavilions designed as an overview of AI excellence at EXPO 2020. Each AI case is showcased on a web board, focusing on different human-centric aspects. In a networking activity during the two days of the EU AI at Expo DUBAI, an AI art performance was presented by Franz Rosati, a musician & digital artist, showing what AI can co-create with artists and how artists can use AI in their creative process.