Friction Is a Feature, an Interview with Bruce Schneier

A companion piece written by William Gao. Interview by Jessica Tang and Doga Baskan.

The Name
May 26, 2026

Friction is a pillar of democracy. Selective friction nurtures pluralism, tempers the more radical impulses of governance, and demands deliberation from all who participate in the dialectic process of public decision-making. Technological innovations often reduce friction, though few have done so as rapidly or as broadly as artificial intelligence (AI). We already see large language models accelerating the pace and efficiency of communication, and autonomous systems executing decisions with an authority once reserved for humans.

What, then, becomes of a democratic system when its strategic slowness is eroded? And conversely, what happens to an authoritarian regime when these obstructions to the accumulation of power are removed? These questions sit at the heart of Rewiring Democracy, where Bruce Schneier examines how AI is reshaping political systems—from elections and lawmaking to courts and law enforcement.

Trained in the mathematics of security, Schneier has developed a distinctive lens on the relationship between technology and power. He has authored 18 books on cryptography, computer security, and policy, and writes the influential newsletter Crypto-Gram and blog Schneier on Security. He is currently spending the 2025-26 academic year at the University of Toronto’s Munk School of Global Affairs & Public Policy, where he leads reading groups and teaches cybersecurity policy.

We had the pleasure of speaking with Schneier about AI governance and policy: recent triumphs, persistent challenges, and its role in engineering the future of capitalism, democracy, and power. The full interview is available on our YouTube channel.

The cost of communication

Friction is not uniformly desirable in democracy. Linguistic barriers, in particular, exclude citizens from political discourse and limit access to government services. AI-powered translation tools directly target this form of friction by lowering the cost of communication across languages and expanding the reach of democratic institutions.

The European Union, bearing 24 official languages, has invested in machine translation to increase the accessibility of its multilingual debates as early as 2019,¹ and today supports a suite of AI-enhanced translation tools.² Similarly, the Government of Canada’s in-house artificial intelligence system GCtranslate reinforces bilingualism in the federal service while exploring support for Indigenous languages.³ In both cases, AI allays linguistic friction between citizens and public services.

Beyond domestic governance, AI amplifies the expressive power of democratic institutions. Since the second day of the war in Ukraine, President Volodymyr Zelenskyy has enlisted Ukrainian AI startup vidby to dub his speeches in more than 20 languages.⁴ By minimizing the cost and delay of translation, AI enables political messages to circulate globally with unprecedented speed and spread.

Across the nations, from public amenities to urgent communications, the decline of linguistic friction following the adoption of AI tools is empowering the voices of citizens wishing to participate as well as authorities seeking to extend their reach.

The unbounded exercise of authority

The preceding proposition features a contrary threat to democratic principles. AI magnifies power where it already exists, rendering many financial and technological constraints toothless against the unbounded exercise of authority. In authoritarian contexts, technologies like biometric recognition and mass data aggregation eliminate the obstacles that normally restrain unchecked control. The People’s Republic of China operates the world’s largest AI-driven public surveillance system, reportedly encompassing seven-hundred million cameras, humanoid robot patrols, and automated systems that monitor and evaluate citizen behaviour.⁵ By replacing human judgment with algorithmic inference, these systems bypass the friction that is integral to democratic establishments.

Closer to home, several agencies under the U.S. Department of Homeland Security, including Immigration and Customs Enforcement, regularly deploy AI in the criminal and civil enforcement of federal law.⁶ The use cases range from email analytics to facial recognition tools. In such high-impact applications, this automation can obscure procedural resistance, allowing biases to propagate and inflating the potential for unruly governance. Somewhat reassuringly, OpenAI’s recent contract with the U.S. Department of War explicitly forbids the use of its technology for mass domestic surveillance,⁷ suggesting that policymakers and technologists alike possess the ability to decide how to implement friction to preserve the integrity of democratic systems.

The Future of Friction

Building on these examples, the question becomes how we can concurrently develop AI technology and policy to decrease friction where doing so dampens democracy and to increase it where doing so protects the public. A cohesive effort to shape the future of friction calls for collaboration between politicians, engineers, and civic stakeholders. As an expert in technology concerned about the social consequences of AI, Bruce Schneier is mediating conversations between technical, philosophical, and political voices with the aim of informing policy design to ensure AI develops in harmony with egalitarian values.

1. Directorate-General for Research and Innovation. (2019). Live Speech to Text and Machine Translation Tool for 24 Languages, EU Funding and Tenders Portal. https://ec.europa.eu/info/funding-tenders/opportunities/portal/screen/opportunities/tender-details/5249.

2. Directorate-General for Translation. AI-Based Multilingual Services. European Commission. https://language-tools.ec.europa.eu/.

3. Public Services and Procurement Canada. (2025). GCtranslate: Using artificial intelligence to build a more agile, modern and bilingual public service. Government of Canada. https://www.canada.ca/en/public-services-procurement/news/2025/09/gctranslate-using-artificial-intelligence-to-build-a-more-agile-modern-and-bilingual-public-service.html.

4. Field, A. (2022). Ukraine Startup Translates Videos For Zelensky, While Adjusting To Work In A War Zone. Forbes. https://www.forbes.com/sites/annefield/2022/07/29/ukraine-startup-translates-videos-for-zelensky-while-adjusting-to-work-in-a-war-zone/.

5. Weber, V. (2025). China’s AI-Powered Surveillance State. Journal of Democracy 36 (4), 151-160. https://dx.doi.org/10.1353/jod.2025.a970356.

6. U.S. Department of Homeland Security. (2026). United States Immigration and Customs Enforcement—AI use cases. https://www.dhs.gov/ai/use-case-inventory/ice.

7. OpenAI. (2026). Our agreement with the Department of War. https://openai.com/index/our-agreement-with-the-department-of-war/.

Categories: Articles, Audio, Recorded Interviews