While concerns about artificial intelligence undermining democratic institutions dominate headlines, a different narrative is emerging across global democracies. From Tokyo courtrooms to California newsrooms, AI is being deployed to strengthen governance, enhance citizen participation and make democracies more responsive.
Japan's AI-Powered Political Revolution
In Japan, a political transformation is underway through innovative AI applications. Engineer Takahiro Anno, then 33, leveraged artificial intelligence to revolutionise his campaign for Tokyo governor. Running as an independent candidate against 55 opponents, Anno deployed an authorised AI avatar that answered an astonishing 8,600 voter questions during a continuous 17-day YouTube livestream.
This groundbreaking approach propelled Anno to fifth place in the crowded field. Two months later, he achieved an even more significant victory, winning election to Japan's upper legislative chamber. This time, his AI systems engaged with over 20,000 constituent queries.
Anno's political party, Team Mirai, now functions as an AI-enabled civic technology organisation. The party is developing the Mirai Assembly app, using Japan's public political funding to create software that enables constituents to express opinions on legislative bills and organise their input using artificial intelligence. Party members have committed to directing their parliamentary questioning based on this AI-processed public feedback.
Brazil's Judicial Transformation Through Automation
Brazil faces a unique challenge with one of the world's most litigious societies, boasting even more lawyers per capita than the United States. The resulting court backlogs have created enormous financial burdens, with the federal government spending approximately 1.6% of GDP annually operating courts and another 2.5-3% of GDP on court-ordered payments from lost lawsuits.
Since 2019, the Brazilian government has aggressively implemented AI throughout its judiciary system. Rather than replacing human judges, AI assists with caseload distribution, legal research, hearing transcription, identifying duplicate filings, preparing initial orders and clustering similar cases for joint consideration.
The results have been dramatic. Brazil's federal supreme court backlog dropped in 2025 to its lowest level in 33 years. However, an interesting development emerged: Brazilian lawyers quickly adopted the same AI tools, leading to a nearly 40% increase in new case filings over five years. Some democracy advocates argue this litigation surge represents healthy civic participation, as citizens use the courts to hold government accountable.
Germany's AI Voting Guidance Innovation
In Europe, Germany has pioneered AI applications to better inform voters. Since 2002, the German Federal Agency for Civic Education has operated Wahl-o-Mat, a non-partisan voting guide. The process involves 24 young voters and education experts developing 80 questions posed to political parties, narrowed to 38 key topics for public consumption.
Recently, outside groups have enhanced this system with AI alternatives. German AI company AIUI created Wahlweise, while students at the Technical University of Munich developed Wahl.chat, an interactive AI system used by more than 150,000 people within four months. These tools allow citizens to engage in conversational AI interactions about party positions rather than reading static webpages.
German researchers studying these tools ahead of the 2025 federal election identified concerns about AI bias and "hallucinations" - systems generating false information. They recommended implementing scientific evaluation standards comparable to the official Wahl-o-Mat to improve reliability while acknowledging the technology's potential to increase voter knowledge and party transparency.
California's AI Watchdog Journalism
In the United States, California demonstrates how AI can strengthen democratic oversight through journalism. Non-profit news organisation CalMatters launched its Digital Democracy project in 2023, collecting every public statement by California elected officials along with voting records, legislation and campaign contributions in a free online platform.
This year, CalMatters introduced AI Tip Sheets, using artificial intelligence to analyse this comprehensive dataset for anomalies. The system identifies patterns such as voting position changes coinciding with large campaign contributions, providing journalists with data-driven story ideas and analysis.
This approach represents AI augmentation rather than replacement of human journalists. The innovation emerged from a non-profit news model, offering technological support to maintain watchdog functions as traditional newspaper business models decline.
The Future of Public AI for Democracy
These global examples share a common theme: AI distributing rather than concentrating power. The technology assists people in performing democratic tasks rather than replacing them entirely. While humans could theoretically perform all these functions, practical limitations make AI assistance invaluable for improving efficiency and facilitating government-public engagement.
A significant barrier involves the dominance of proprietary AI systems developed by US tech giants. In response, advocates promote "public AI" - models developed under democratic control for public benefit. Switzerland recently released Apertus, the world's most powerful fully realised public AI model, developed jointly by the Swiss government and ETH Zurich university.
Apertus represents a breakthrough as an entirely open-source system free from illegally acquired copyrighted training materials and exploitation of global south labour. While its performance matches corporate AI from approximately one year ago, it demonstrates that trillion-dollar investments aren't necessary to create effective AI models.
Artificial intelligence inherently amplifies power, capable of enhancing both authoritarianism and democracy. The direction depends on human choices. As more citizen watchdogs, litigators and political parties use AI to amplify oversight capabilities and engage voters, while governments provide democratic alternatives to corporate AI, society stands to benefit significantly from this technological transformation.
