Published on February 19, 2026
This week on Reboot Democracy, Max Stier explores how AI can strengthen internal government culture by building real-time recognition systems for civil servants. Elana Banin examines the UbuntuGuard benchmark, arguing that AI safety must be tested against locally defined public-sector rules rather than inferred from English-language standards. Giulio Quaggiotto reflects on Argentina’s AI-supported “Questions Tree” experiment to rethink how governments build strategy. Beyond Reboot, Colombia’s AI-presented candidate Gaitana enters an Indigenous election, the Pentagon pressures Anthropic to loosen military guardrails, global experts release the 2026 International AI Safety Report, governments expand AI-assisted legislative drafting, and India convenes the first major Global South AI summit.
InnovateUS delivers no-cost, at-your-own-pace, and live learning on data, digital, innovation, and AI skills. Designed for civic and public sector, programs are free and open to all.
Using AI to Understand Public Sentiment – February 23, 2:00 PM ET
Introducing Grantwell: Using AI for Municipal Technical Assistance – February 25, 2:00 PM ET
Effective Use of Social Media: Storytelling, Trust, and Institutional Brand – February 26, 2:00 PM ET
Prompting Lab Office Hours: Bring your AI Questions – February 27, 2:00 PM ET
AI and Elections
Colombia’s March 8 Indigenous special-district elections will feature “Gaitana,” an AI-presented candidate advocating digital sovereignty and participatory democracy. While marketed as an AI congresswoman working 24/7 without a salary, election authorities clarified that a human candidate is formally registered, campaigning via an AI interface labeled “IA” on the ballot.
Read articleGoverning AI
The Pentagon is reportedly considering severing ties with Anthropic after the company refused to authorize unrestricted military use of Claude, particularly for mass domestic surveillance or fully autonomous weapons. OpenAI, Google, and xAI have reportedly shown greater flexibility. The dispute underscores a structural tension: commercial AI labs’ usage policies versus defense agencies’ demand for “all lawful purposes” access.
Read articleGoverning AI
Backed by 100+ experts from 30+ countries, the second International AI Safety Report assesses frontier AI capabilities, risks, and technical safeguards. It organizes risks into misuse, malfunctions, and systemic harms, while highlighting the “evidence dilemma” policymakers face as capabilities outpace empirical risk data. The report stops short of policy prescriptions, instead offering a shared scientific baseline for governments navigating AI governance.
Read articleGoverning AI
Adapting the classic internet governance model, the authors map AI governance across infrastructure, logical (models), and social (applications) layers. The framework reveals fragmentation and duplication across institutions while highlighting how frontier firms are vertically integrating across all layers. It’s a useful taxonomy for understanding where governance gaps and power consolidation are emerging.
Read articleGoverning AI
UbuntuGuard tests whether AI systems comply with locally defined policies across ten African languages and six countries. Rather than relying on English-language benchmarks, the framework evaluates alignment with real institutional norms. The findings suggest that AI safety must be policy-explicit and context-specific and that governments need the capacity to test systems against their own rules before deployment.
Read articleAI for Governance
From the UK to Brazil and New Zealand, governments are using AI tools to summarize consultations, cluster amendments, and draft legislative materials. While promising efficiency gains, researchers warn that AI-assisted rulemaking could expose regulations to legal challenge, reduce transparency, or weaken legitimacy if procedural safeguards aren’t preserved.
Read articleAI for Governance
POPVOX released practical AI toolkits for legislative offices, including use policies, change management guides, custom GPT instructions, and deep research workflows. The effort reflects a broader push to address the “pacing problem” by equipping parliaments with operational AI fluency rather than abstract literacy alone.
Read articleAI for Governance
Drawing on federal workforce data, Max Stier argues that rebuilding trust in government begins internally. AI can help automate recognition programs, reduce bias in performance acknowledgment, and personalize praise, strengthening morale and outcomes. The piece reframes AI not just as efficiency infrastructure but as cultural infrastructure.
Read articleAI for Governance
Argentina’s Red de Innovación Local piloted an AI-supported “Questions Tree” strategy process using PortalRIL. By surfacing patterns and trade-offs across teams, AI compressed months of coordination into weeks. The experiment suggests AI’s real promise lies not in automation but in expanding collective strategic insight.
Read articleAI for Governance
India hosted the first major AI summit in the Global South, convening heads of state and tech executives under themes of People, Planet, and Progress. The summit positioned India as a bridge between frontier AI development and Global South implementation needs, while unveiling 12 indigenous foundation models.
Read articleAI and Public Safety
The MTA is piloting AI-enabled fare gates that detect and document fare evasion through camera-triggered clips and AI-generated descriptions. Vendors are competing for a $1.1B modernization contract. The rollout illustrates how AI-driven monitoring is embedding into civic infrastructure.
Read articleAI and Public Safety
Police departments are purchasing GeoSpy, an AI tool that infers where photos were taken using environmental visual cues. While pitched as lead-generation support, the reporting raises questions about validation, error rates, and mission creep in AI-assisted geolocation.
Read article