Artificial Intelligence, Big data and the Internet of Things - report from day 3 IGF 2018, 14.11.2018Jutta Croll, Stiftung Digitale Chancen
Day three of the IGF started with the Best Practice Forum on Artificial Intelligence. Such thematic forums are part of the activities that take place between the annual IGFs, with the intention of exploring topics of current interest in greater depth.
Nobuhisa Nishigata, OECD pointed out how difficult it is to exploit the benefits of artificial intelligence while keeping risks under control. Imane Bello, Sciences Lecturer, Human Rights & AI, proposed to avoid the terms Artificial Intelligence and the Internet of Things, after all, it's all about people, not just machines, she said
Taylor Bentley, ISED, Canadian Government, explained the Canadian approach of working with business and academia to develop legislation for IoT, AI and Big Data. However, there are currently more examples of unsafe and risky applications, the representative of MISEC, an IT security company, said.
The difficult balance between security and innovation was also addressed by Michael Nelson, Cloudflare. All participants agreed that IoT, AI and Big Data offer excellent opportunities to make the world a better place. But this can only succeed if human rights are respected and the needs of particularly vulnerable groups are taken into account and the risks minimized. This requires a balanced relationship between government regulation and the willingness of companies to implement Safety by Design as a principle of development and design.
Human Rights and the impact of domain names was the topic of another session this morning. Human Rights Impact Assessments (HRIAs) are a useful and increasingly widespread tool to inform private actors of the potential negative impacts of their policies, and to help mitigate their consequences. Currently, there is little mutual acceptance of the results of HRIAs if they were carried out either by business or civil society, says Michelle Neylon, representative of an Irish company responsible for domain name registration. The proposal to implement ICANN's and IGF's multi-stakeholder model for HRIAs could help avoid misunderstandings. However, companies are generally sceptical about the enforcement of human rights with regard to additional contractual requirements. In response to questions, those involved in the session stressed that child sexual abuse representations are content that violates human rights. There was a willingness to tackle the problem together, but further discussion of contractual and technical possibilities was needed. At the beginning of December, an article will appear in the magazine Medien + Erziehung - merz Wissenschaft, pointing out a way to do this. Access to the English translation of the text will be published here.