Safeguarding freedom of expression in the age of AI

Safeguarding freedom of expression in the age of artificial intelligence

Rome/Vienna [ENA] On 26 January 2022, OSCE Representative on Freedom of the Media Teresa Ribeiro launched the SAIFE Policy Manual, a publication that puts a spotlight on the impact of artificial intelligence on freedom of expression. The OSCE Representative on Freedom of the Media observes media developments in all 57 OSCE participating States. She provides early warning on violations of freedom of expression and media freedom and

promotes full compliance with OSCE media freedom commitments. “The way online information is curated and moderated has a direct and significant impact on global peace, stability and comprehensive security,” Ribeiro said. “This year’s 25th anniversary of the Mandate of the OSCE Representative on Freedom of the Media provides an important moment in time to address the challenges to freedom of expression that stem from the use of digital technologies, including automation and artificial intelligence (AI).”

The SAIFE Policy Manual is the culmination of two years of research and several expert workshops, bringing together the knowledge and expertise of more than 120 of the most renowned scholars and practitioners in the field of freedom of expression, media freedom, technology, and security. The publication provides comprehensive policy guidance for States to ensure that online information spaces are in line with international human rights standards and realize the key principles of transparency, accountability and public oversight. The SAIFE Policy Manual – which is the outcome of a successful partnership between the OSCE and the digital rights organization Access Now – provides technical, actionable and human rights-centric recommendations

on safeguarding free speech while using AI in content governance. In particular, it focuses on the use of AI in content moderation to detect and evaluate (a) illegal content online, including security threats such as extremist and terrorist content, and (b) potentially harmful but legal content, with a specific focus on online hate speech and algorithmic discriminatory bias. The Manual also focuses on AI that is designed and deployed for curating and personalizing online content and its impact on media pluralism, as well as its nexus to surveillance-based advertising. The SAIFE Policy Manual can be found here:  https://www.osce.org/representative-on-freedom-of-media/510332