Dario Amodei envisions millions of autonomous AI agents as economic actors [1] and warns of power concentration as an existential risk [2], yet his responses remain in the domain of policy. Policy governs from the outside. What's missing is infrastructure that governs from within: rules designed so that self-interested autonomous behavior produces collectively good outcomes. This perspective argues that mechanism design provides a missing conceptual lens for understanding and governing the emerging agentic economy.
The technical building blocks for such systems are already emerging. Machine-native payments through x402 [3], on-chain agent identity and reputation in the direction of ERC-8004 [4][5], smart contracts as enforceable commitments. These are not isolated tools. Taken together, they form the primitives of a global, tamper-resistant institutional layer for autonomous agents. Yet infrastructure alone is not sufficient. Designing the rules that autonomous agents play by (verification, incentives, and reputation) is fundamentally a mechanism design problem [6].
The challenge of designing incentive-compatible rules for interacting autonomous agents becomes even more pressing as AI systems increasingly acquire physical embodiments [7][8][9]. Consider autonomous vehicles operated by different companies sharing road infrastructure with human drivers and delivery drones. Different operators, different jurisdictions, no shared authority. A vehicle could report an obstacle that does not exist to claim priority routing. Without verifiable telemetry and incentive-aligned reporting, shared situational awareness degrades and safety is compromised. How do you make truthful reporting the dominant strategy? How do you design rules so that self-interested autonomous machines cooperate reliably in shared physical space?
While a misbehaving software agent may cause financial loss, an embodied agent might cause harm. Regulatory frameworks designed for human actors may struggle to scale to such environments [10][11]. Blockchain-based systems offer a potential substrate where institutional rules for the machine economy can be encoded and enforced. How these rules are designed will play a central role in determining how power, coordination, and accountability emerge in the agentic economy. It also raises a deeper question: mechanism design was built for rational actors, but AI agents are a new class of participant, one that optimizes objectives consistently yet can also behave unpredictably. Whether our existing frameworks apply, or whether we need new ones, is an open question that deserves serious attention.
Attendees will leave with a new perspective on the agentic economy and the problem of designing rules for autonomous actors, along with a set of open questions about whether existing mechanism design tools are sufficient for this new class of participant.

References
[1] D. Amodei, Machines of Loving Grace, October 2024. https://www.darioamodei.com/essay/machines-of-loving-grace
[2] D. Amodei, The Adolescence of Technology, January 2026. https://www.darioamodei.com/essay/the-adolescence-of-technology
[3] x402 Protocol (Coinbase / Cloudflare, 2025). Stripe integration Feb 2026. https://x402.org
[4] ERC-8004: Trustless Agents (Draft). Crapis, De Rossi, Ellis, Reppel, Aug 2025. https://eips.ethereum.org/EIPS/eip-8004
[5] Ethereum Foundation dAI Team, Sep 2025. https://www.coindesk.com/tech/2025/09/15/ethereum-foundation-launches-ai-team-to-support-agentic-payments
[6] P. Rawson / Ecofrontiers, AI Mechanism Designer: The Job That Doesn't Exist Yet. https://www.linkedin.com/pulse/ai-mechanism-designer-job-doesnt-exist-yet-ecofrontiers-0f1ye/
[7] Deloitte, Physical AI and Humanoid Robots, Tech Trends 2026. https://www.deloitte.com/us/en/insights/topics/technology-management/tech-trends/2026/physical-ai-humanoid-robots.html
[8] NexGen Architects, 9 Predictions of Agentic AI in 2026. https://www.nexgenarchitects.com/blog-posts/agentic-ai-predictions-2026
[9] NVIDIA Physical AI announcements (2025-2026): Cosmos models, Isaac GR00T N1.6. https://developer.nvidia.com/isaac/gr00t
[10] Osborne Clarke, Robotics at a Global Regulatory Crossroads, Apr 2025. https://www.osborneclarke.com/insights/robotics-global-regulatory-crossroads-compliance-challenges-autonomous-systems
[11] EU AI Act (Regulation EU 2024/1689), high-risk rules effective Aug 2026. https://artificialintelligenceact.eu