The rise of artificial intelligence (AI) presents both opportunities and risks to freedom of expression in the 21st century. Creative AI serves as a tool to enhance thinking, writing, searching, and creating, proving invaluable for dissidents in authoritarian regimes by circumventing censorship and exposing suppressed information.

However, the emergence of intrusive AI poses a significant threat. Authoritarian governments are employing real-time surveillance, facial recognition, and censorship, all powered by AI, to control populations.

Even AI models operated by private companies demonstrate issues. Testing reveals that chatbots frequently refuse to answer legal but controversial questions, raising concerns about uniformity and access to information controlled by a few tech giants. The concentration of power in the hands of a few companies allows for the filtering of information based on political and commercial interests.

Two main approaches to AI regulation are emerging: user responsibility and preventive control. The EU’s AI regulation attempts to balance these, but risks overregulation due to imprecise requirements.

Taiwan offers a third approach, employing open, citizen-driven solutions. AI is used for crowdsourced fact-checking and digital citizens’ assemblies, strengthening democracy without restricting freedom of expression.

Ultimately, the path chosen in navigating the opportunities and risks of AI will determine the future of freedom of thought and expression. Defending the open society while using technology wisely is paramount.