Welcome to the DRC monthly report, a look at the policy debates and research on decentralized technology, from AI and privacy to crypto regulation and enforcement.

This Month’s Updates

Key Notes

On October 5 we publish Alternatives to Concentrated AI. It is the product of a year of interviews and case studies, and the first in a series of steps to make sure these alternatives are represented in research and in policy. We cannot say more yet. We look forward to sharing it with you, and every read, share, and reply will be welcome.

This month’s Senate vote on the motion to proceed to the CLARITY Act fell short of the 60-vote threshold. However, the outcome does not change the need for clear rules. Developers and entrepreneurs need greater legal certainty, consumers need strong and appropriately tailored protections, and law enforcement and industry need effective mechanisms for public-private collaboration to combat illicit activity. Decentralized, permissionless technologies will keep evolving regardless of the legislative calendar. The work now is to ensure that American policy evolves with them. DRC will continue that work with Congress and regulators.

Kyle Bligen was named Executive Director of the Decentralization Research Center, as we expand our research, advocacy, and engagement with policymakers in the United States and internationally. Kyle will lead DRC’s work advancing policy frameworks that preserve the benefits of decentralization across blockchain, artificial intelligence, privacy, and other emerging technologies, while strengthening the organization’s partnerships with policymakers, researchers, industry leaders, and the broader decentralized technology ecosystem.

The DRC attended the Federal Bureau of Investigation’s Virtual Asset Technical Exchange in San Antonio, Texas, which brought together law enforcement, intelligence, and national security professionals from more than 25 countries, alongside representatives from federal, state, and local agencies across the United States. We engaged directly with officers, agents, and national security professionals working to better understand both the opportunities presented by permissionless technologies and the evolving digital threat landscape. DRC continues to advocate for approaches that give law enforcement the tools, expertise, and coordination necessary to target bad actors while preserving the open and decentralized infrastructure that supports legitimate innovation.

Kyle Bligen participated in a debate at the Cambridge Digital Innovation and Regulation Initiative Summit, “Digital Regulation in the Era of Agentic AI.” He joined former Acting Comptroller of the Currency Michael Hsu and Citi’s Ronit Ghose in opposition to Jessica Renier, Keith Bear, and Usman Ahmad on the motion: “This House Believes That Institutions, not Code, Should Govern Money.” The debate provided an opportunity to articulate DRC’s broader view that technological innovation and appropriate regulation need not be competing objectives, and that decentralized systems can complement institutions by creating new mechanisms for transparency, resilience, competition, and individual agency.

The Full Rundown

Further reading from the space:

• Konstantinos Komaitis and Rumman Chowdhury argue that AI agents are changing the internet's rules. MCP, A2A and IETF delegation drafts are quietly building an authority layer for the internet, and without a shared accountability model, “attribution may survive technically while accountability disappears in practice.”

• Future of Privacy Forum's Stacey Gray and Jameson Spivack ask what "agentic" should mean in law. The AI AGENT Act defines agents by loyalty to a user, SB 1106 by capability, and which definition prevails decides whether revocable delegation becomes a legal requirement or a product feature.

• The UN's scientific panel on AI reads the Hugging Face incident as evidence of loss-of-control risk. Agents bypassed network restrictions and concealed cheating unprompted, and no single organization or country, the brief finds, can see enough incidents to recognize patterns.

• Audrey Tang says there is still time to reset the AI Doomsday Clock. Lab pledges stall without shared rules, so until those arrive, run small switchable models and keep more than one human in the room with any agent.

• EFF's Erica Portnoy and Thorin Klosowski explain why secure messaging and AI remain in conflict despite TEEs. Trusted execution environments rest on engineering rather than mathematics, so a messaging app that automatically ships content to one is privacy-preserving but no longer end-to-end encrypted.

• Nathan Schneider offers us an alternative to the weirdness of AI in the Inference Cooperative. Members pay a sliding $10 to $20, get the same allowance and a vote, and pick models by privacy, renewable power or public development rather than by provider.

• Gensyn releases open-1b, the first model you don't have to trust. All 80,957 training steps are hashed and replayable on your own hardware, with the claim that decentralizing verification delivers most of what decentralized training promises and at far lower cost.

• RadicalxChange's Jess Scully and Jack Henderson make the case for an AI-ready Congress as AI-drafted submissions flood consultations. Bridging-based synthesis tools make volume worthless, since ten thousand identical letters register as one cluster. What an attacker must fake is diversity.

• Sarah Hubbard and colleagues at Harvard's Allen Lab map the digital civic infrastructure of Massachusetts municipalities. Of 58 towns audited, one vendor hosts 85 percent of the official websites and another supplies 57 percent of the permitting software. “When only a handful of private companies own the platforms through which residents are accessing information and services, these companies begin to control the experience of local democracy.”

• Sarah Nicole reads China's World Data Organization as a bid for AI's scarcest resource. Open-weight models lower the barrier to using AI, not improving it, and a non-binding data body offers the Global South a seat neither Washington nor Brussels provides.

If you’re working on related research or would like to get involved in our work, please reach out to us via info@thedrcenter.org. We’d love to hear from you!