Policy Manual

This publication is part of the project “Spotlight on Artificial Intelligence and Freedom of Expression” (#SAIFE).

Introduction: Upholding the Principles of the Helsinki Final Act in the Digital Age

In the framework of the SAIFE project, the Office of the OSCE Representative on Freedom of Expression, together with Access Now, organised four expert workshops in the first half of 2021. These expert workshops unpacked and analysed the main challenges that AI tools pose to human rights, in particular, the right to freedom of expression and opinion, and media freedom and pluralism. The workshops focused on four main thematic issues:

This report contains the main findings of these expert workshops as well as policy recommendations addressed to the OSCE participating States, while acknowledging that a multi-stakeholder approach is needed to effectively and sustainably address the complex challenges that content moderation and content curation pose to freedom of expression. The recommendations for OSCE participating States were put forward during the workshops and reviewed by renowned experts in the field of freedom of expression, media pluralism and artificial intelligence.

Read full publication

Key Recommendations for OSCE Participating States

  1. Protect and promote freedom of expression and other human rights as the centre of AI-related strategies and policies
  2. Preserve and foster the internet as a space for democratic participation and representation and for media pluralism
  3. Develop evidence-based policies, built on inclusive processes, to respond to challenges to freedom of opinion, freedom of information and freedom of expression
  4. Promote compliance with the UN Guiding Principles on Business and Human Rights, to prevent the prioritisation of profit maximisation at the expense of human rights and democratic values
  5. Oblige online platforms to conduct human rights due diligence, including through human rights impact assessments (HRIAs) for their content governance policies and automated decision- making, as well as for their business practices, such as data harvesting, targeted advertising and interface design
  6. Enforce clarity, explainability, and accessibility on the use of AI for content moderation, content curation and targeted advertising
  7. Ensure that human rights protections are not fully outsourced or automated, and provide transparency about any public-private- partnerships
  8. Enact strong transparency frameworks, including by mandating comprehensive transparency reports that contain detailed information on the use of AI
  9. Make certain that robust remedy mechanisms against censorship and surveillance power are in place, including through human review and independent appeal mechanisms
  10. Guarantee strong accountability, including through independent oversight and independent auditing, particularly of compliance with human rights and non-discrimination
  11. Respect the right to privacy and data protection, including by identifying limits to surveillance-based advertising and by ensuring robust transparency and user agency in tracking- and profiling-based business practices
  12. Promote media and digital literacy and strengthen users' empowerment, agency and control over content governance and the use of their data, including by providing the possibility to opt-out of all automated decision-making
  13. Address unbalanced and monopolised market powers and promote plurality, and technological and media innovation
  14. Engage on the multilateral level to ensure human rights s afeguards in the development and deployment of AI for online content curation and content moderation

EXPERTS

Authors

Eliska Pirkova, Matthias Kettemann, Marlena Wisniak, Martin Scheinin, Emmi Bevensee, Katie Pentney, Lorna Woods, Lucien Heitz, Bojana Kostic, Krisztina Rozgonyi, Holli Sargeant, Julia Haas, and Vladan Joler

Editors

Deniz Wagner and Julia Haas

Experts

Jennifer Adams, Susie Alegre, Asha Allen, Andreas Marckmann Andreassen, Nikolett Aszodi, Jef Ausloos, Josephine Ballon, Joan Barata, Nadia Bellardi, Susan Benesch, Guy Berger, Frederik Zuiderveen Borghesius, Irina Borogan, Jonathan Bright, Elda Brogi, Amy Brouillette, Joanna J. Bryson, Pete Burnap, Camilla Bustani, Ignacio Talegon Campoamor, Maja Cappello, Marcelo Daher, Anita Danka, Nicholas Diakopoulos, Aijamal Djanybaeva, Leyla Dogruel, Maria Donde, Sead Dzigal, Franccesca Fanucci, Marc Fumagalli, Maximilian Gahntz, Jana Gajdošová, Maya Indira Ganesh, Lalya Gaye, Brandi Geurkink, Arzu Geybulla, Michele Gilman, Nadine Gogu, Gabrielle Guillemin, Rustam Gulov, Ben Hayes, Natali Helberger, Georgia Holmer, Andrea Huber, Karolina Iwańska, Sam Jeffers, Elliot Jones, Pascal Jürgens, Agnes Kaarlep, Frederike Kaltheuner, Kari Karppinen, Susan Kerr, Benjamin Kille, Yoojin Kim, Wolfgang Kleinwächter, Beata Klimkiewicz, Djordje Krivokapić, Ľuboš Kukliš, Andrey Kuleshov, Joanna Kulesza, Collin Kurre, Susan Landau, Paddy Leerssen, Emma Llansó, James MacLaren, João Carlos Magalhães, Samvel Martirosyan, Estelle Massé, Kyle Matthew, Eleonora Maria Mazzoli, Tarlach McGonagle, Marko Milosavljevič, Mira Milosevic, Iva Nenadić, Marielza Oliveira, Rebekah Overdorf, Roya Pakzad, Sejal Parmar, Patrick Penninckx, Jon Penny, Carlos Perez-Maestro, Emilija Petreska-Kamenjarova, Andrej Petrovski, Courtney Radsch, Otabek Rashidov, Judith Rauhofer, David Reichel, Moritz Riesewieck, Katitza Rodriguez, Asja Rokša-Zubčević, Bianca Schönberger, Christopher Schwartz, Lisa Seidl, Murtaza Shaikh, Jat Singh, Vanja Škorić, Andrei Soldatov, Maria Luisa Stasi, Nikolas Suzor, Damian Tambini, Dhanaraj Thakur, Gulnura Toralieva, Max van Drunen, Vitaly Vasilchenko, Francisco Vera, Kristina Voko, Diana Vlad Calcic, Ben Wagner, Douglas Wake, Hilary Watson, Agnieszka Wawrzyk, Veszna Wessenauer, and Andrej Zwitter

Observers

United Nations OHCHR, UNESCO, Council of Europe, European Audiovisual Observatory, European Commission, European Union Agency for Fundamental Rights, European Broadcasting Union, OSCE (Secretariat, ODIHR, HCNM)