The 13th annual Freedom on the Net report by Freedom House reveals a concerning decline in global internet freedoms, posing potential worsening threats with the use of artificial intelligence tools in undemocratic ways. The report, spanning from June 2022 to May 2023, highlights troubling records, including 55 countries where individuals faced legal repercussions for their online speech and 41 countries restricting access to websites hosting political and social content.
China retains its position as the worst offender in internet restrictions for the ninth consecutive year, with Myanmar as the second-worst, aggressively targeting critics of the 2021 military coup and reinforcing state censorship.
The report underscores the significant potential for AI to harm digital rights if misused. It emphasizes AI’s capacity for both good and harm, with generative AI being used in influence operations across 16 countries.
AI-generated content, including manipulated images of political figures, has been disseminated globally. Authoritarian governments in some countries are attempting to control and utilize generative AI for censorship. China, for example, prohibited tech giants Tencent and Ant Group from integrating ChatGPT, while Apple removed similar apps from its Chinese app store.
AI is also employed for content moderation on social media platforms, often under government orders in 22 countries. While AI has the potential to efficiently detect and remove banned speech, the report highlights that in many parts of the world, it is used as another tool in the authoritarian arsenal. To address these challenges, regulations are deemed essential, with the proposed European Union Artificial Intelligence Act expected to pass later this year.
When deployed safely and fairly, AI can help counter authoritarian censorship, combat disinformation, and document human rights abuses, providing a promising solution for the future of internet freedoms.