Schneier's June 2025 testimony before the House Committee on Oversight and Government Reform — "The Federal Government in the Age of Artificial Intelligence" — represents his most direct and recent statement to Congress on AI governance. It is a companion piece to rewiring-democracy (forthcoming at time of testimony), bringing the book's arguments into the legislative hearing context.
The Policy Context
The testimony came at a period of intense congressional attention to AI's implications for government operations, national security, and democratic governance. The federal government was actively deploying AI systems in regulatory, law enforcement, and administrative contexts, while simultaneously confronting AI-enabled threats to elections, public discourse, and critical infrastructure. The Oversight and Government Reform Committee's jurisdiction over federal operations made it a natural venue for questions about how the government should both deploy AI responsibly and respond to AI-enabled risks.
The Argument
Schneier's testimony structured the AI challenge through his established analytical frameworks:
The trust-framework from liars-and-outliers provided the diagnostic: AI disrupts the mechanisms by which democratic societies maintain trust at scale. When AI can fabricate evidence, impersonate officials, automate the generation of disinformation, and personalize manipulation at individual scale, the societal trust infrastructure faces stresses for which it was not designed.
The security-mindset provided the analytical orientation: the question is not whether AI has impressive capabilities but what the attack surface looks like. How will adversaries — foreign and domestic — use AI to exploit vulnerabilities in democratic institutions? What are the gaps, ambiguities, and unintended affordances in the systems governing elections, public discourse, and government administration that AI makes newly exploitable?
hacking-as-systems-subversion provided the power analysis: AI is primarily a tool that amplifies the capabilities of those who already have resources. Governments with large AI programs, corporations with data and compute, and well-funded political actors will gain capabilities to hack democratic rule systems that ordinary citizens and small institutions cannot match or defend against.
Policy Recommendations
The testimony advocated for transparency requirements on government AI deployments, procurement standards that include security and fairness audits, and legal frameworks establishing government liability for AI-enabled harms. Reflecting the evolution from earlier testimonies, it was more prescriptive and more explicit about the political economy of AI governance — who benefits from weak regulation and why.
Connection to the Schneier-Sanders Collaboration
The testimony draws directly on work developed with nathan-sanders for rewiring-democracy. The combination of Schneier's security analysis and Sanders' data science expertise gives the policy arguments a technical grounding that distinguishes them from the more speculative AI policy commentary that was common in 2025.