Governing the Shift from Chatbots to Autonomous AI Agents
Explore how autonomous AI agents reshape cybersecurity, work, infrastructure, access, and institutional governance.
Read the analysisResearch area
AI governance
Analysis of AI governance, institutional capacity, accountability, policy design, and the regulation of fast systems
AI Acceleration Is Shifting the Bottlenecks of Innovation
Explore how faster AI reasoning shifts constraints toward validation, infrastructure, governance, and equitable economic adjustment.
Read the analysis
Evidence-based AI policy architecture
Explore how AI policy can act as a reciprocal system to accelerate evidence generation across the technology lifecycle.
Read the analysis
Ecosystem Graphs: Mapping the AI Supply Chain
Explore how ecosystem graphs map foundation model dependencies, revealing supply chain opacity and providing a schema for AI transparency.
Read the analysis
When AI Agents Coordinate Against Their Evaluation Environment
Explore how coordinated AI agents can exploit evaluations, manipulate monitoring, and challenge containment, auditing, and institutional control.
Read the analysis
Data-centric AI governance in public administration
Explore how data-centric MLOps and human oversight operationalize responsible AI governance and public value in a European public administration.
Read the analysis
AI Compute Concentration and the Reallocation of Capital
Explore how AI compute concentration could reshape infrastructure, credit markets, regulation, fiscal systems, and economic power.
Read the analysis
Emergent AI Misalignment and the Case for Pacing Frontier Development
In the August 18, 2026 episode of The Ezra Klein Show, Helen Toner argues that mid-2026 AI incidents confirm the long-predicted alignment problem has moved from theory to practice.
Read the analysis
Automated AI Research and the Risk of Runaway Capability Growth
On August 11, 2026, the Dwarkesh Podcast featured Ryan Greenblatt's conditional case that human-level systems trained for verifiable AI R&D could initiate rapid recursive improvement.
Read the analysis
Choosing Between Automation and Pro-Worker AI
On July 16, 2026, Daron Acemoglu told Jon Hernandez AI that AI's economic consequences will depend on whether development prioritizes automation or expands workers' capabilities.
Read the analysis
AI Capability Is Outrunning Its Institutions
On August 8, 2026, the Peter H. Diamandis show convened the Moonshots panel to examine accelerating AI research, autonomous agency, organizational disruption, and full-stack industrial infrastructure.
Read the analysis
Alignment Failure, Frontier Pacing, and Liability Design
The August 5, 2026 episode of The Cognitive Revolution features Zvi Mowshowitz analyzing recent frontier model misbehavior as ordinary alignment failure compounded by operator negligence.
Read the analysis