Freedom on the Net 2023: The Repressive Power of Artificial Intelligence

Global internet freedom faced a continued decline, marking the 13th consecutive year of deterioration. Iran witnessed the most severe decline, intensifying digital repression by shutting down internet services, blocking WhatsApp and Instagram, and increasing surveillance to suppress antigovernment protests. Myanmar came close to surpassing China as the world's worst environment for internet freedom, a title China retained for the ninth consecutive year. The Philippines experienced deteriorating conditions, with President Rodrigo Duterte leveraging an antiterrorism law to silence critical news sites. Costa Rica's standing as an advocate for internet freedom was jeopardized after a president's election linked to online trolls harassing major media outlets.

Instances of attacks on free expression escalated globally. In a record 55 out of 70 countries covered by Freedom on the Net, individuals faced legal consequences for expressing themselves online. Additionally, people were physically assaulted or killed for their online commentary in 41 countries, notably Myanmar and Iran, where authoritarian regimes imposed death sentences related to online expression-related crimes. Belarus and Nicaragua witnessed a sharp decline in protections for internet freedom, resulting in draconian prison terms for online speech.

Generative artificial intelligence (AI) poses a significant threat by fueling disinformation campaigns online. At least 47 governments strategically deployed commentators to manipulate online discussions in their favor. Meanwhile, AI-based tools capable of generating text, audio, and imagery have become more sophisticated and accessible, contributing to an alarming escalation in disinformation tactics. In the past year, this technology was employed in 16 countries to sow doubt, discredit opponents, or influence public debate.

Governments have leveraged AI to refine online censorship methods. Advanced authoritarian regimes have embraced AI chatbot technology to enhance their censorship systems, either ensuring compliance or strengthening existing measures. Legal frameworks in 21 countries mandate or incentivize digital platforms to use machine learning to censor undesirable political, social, and religious speech. However, AI hasn't entirely replaced traditional methods of information control; a record 41 governments blocked websites featuring content protected under international human rights law, challenging free expression standards. Even in democratic settings like the United States and Europe, governments considered or imposed restrictions on accessing major websites and social media platforms as a misguided response to foreign interference, disinformation, and online safety.

To safeguard internet freedom, proponents of democracy must adapt insights from past internet governance challenges and apply them to AI. AI can exacerbate digital repression by facilitating censorship, surveillance, and the creation and dissemination of disinformation, making them easier, faster, cheaper, and more effective. An overreliance on self-regulation by private companies has left people's rights vulnerable in the digital age. A reduction in resources within the tech sector could worsen this deficiency. To preserve a free and open internet, democratic policymakers should collaborate with global civil society experts, establishing robust human rights-based standards for both state and non-state actors involved in the development or deployment of AI tools.