The UN Wants to Ban Killer Robots: Why Blockchain Is the Only Way to Enforce It

At the AI Summit in Geneva, UN Secretary-General António Guterres issued what may be the most consequential warning of the decade: "We cannot vibe code the future of humanity." His target was not the latest LLM or generative art craze, but the silent, accelerating race to deploy lethal autonomous weapons systems (LAWS). For those of us who have spent years navigating the intersection of code and trust, his words carried a deeper resonance—they were an indictment of a technocratic culture that believes speed and innovation can outrun moral responsibility. But here is the uncomfortable truth: banning killer robots without a technical framework for accountability is like passing a law against theft without having police or courts. This is where blockchain enters the picture—not as a magical bullet, but as the only infrastructure capable of providing the verifiable, immutable audit trail that meaningful human control requires. The debate over LAWS is not new. Academics and NGOs have warned for years about the responsibility gap—the inability to assign blame when an autonomous system makes a lethal error. Current military systems rely on opaque black boxes; decisions are made in isolated servers. No one can prove afterward whether a human gave the final command or the algorithm acted on its own. This is precisely the kind of informational asymmetry that blockchain was designed to solve. Consider the parallels: the same cryptographic primitives that secure Bitcoin's ledger—hashing, consensus, immutable timestamps—can provide a tamper-proof record of every decision leading up to a strike. Every token holds a story waiting to be mined. In this case, the token is the decision log; the story is the chain of accountability. The soul of the chain is written in its holders—here, the holders are the human decision-makers. We do not just trade assets; we curate narratives. The narrative of 'human control' can be encoded as deterministic logic. Let me ground this in my own experience. During the 2022 bear market, I retreated to a cabin in the Pyrenees to study the economic incentives of Uniswap and Compound. That solitude taught me something deeper: trust in code is only as strong as the audit trail behind it. Later, I audited the smart contracts of several 'AI-driven' protocols. The pattern was consistent: the code claimed to be autonomous, but the owners held backdoor keys. The narrative of decentralization was a facade. The same risk applies to military AI. A blockchain-based system could enforce that no lethal action is taken without a multi-signature approval from at least three verified human officers, each cryptographically signing their intent. Furthermore, smart contracts can be programmed to require a 'cooling-off' period between target identification and engagement, allowing for manual override. This is not science fiction; it is on-chain governance applied to the ultimate use case. In my work with Optimism's RetroPGF transparency, I saw how on-chain allocation reduced nepotism. The same principle can ensure that military orders are transparent to a community of auditors—perhaps international observers—without revealing operational secrets via zero-knowledge proofs. The machinery exists; we need the will to implement it. But there is a dark side that few discuss. A blockchain that records every military decision could just as easily be used by a repressive regime to create an unalterable ledger of 'approved' atrocities, laundering responsibility through code. The technology is neutral; it amplifies the ethics of its designers. Moreover, the latency and bandwidth requirements of consensus mechanisms may be incompatible with real-time combat. A fully permissioned chain overseen by a consortium of nations might work, but that requires geopolitical trust we currently lack. The greatest risk is that blockchain becomes a fig leaf—allowing nations to claim 'accountability' while retaining full control. We must avoid the trap of techno-solutionism. Rather, we need to embed these systems within a robust legal and ethical framework that includes meaningful human review. The contrarian truth is that blockchain alone cannot fix the deeper issue: the lack of political will to cede control to transparent systems. As long as powerful states view LAWS as a strategic advantage, they will resist any binding agreement—crypto or not. Guterres is right to demand a ban on killer robots. But a ban without verification is an invitation to cheat. The future of AI governance will not be decided by resolutions alone—it will be built, line by line, on ledgers that cannot be rewritten. The question is: will we curate a narrative of shared accountability, or will we let the story be written by the fastest coder in the room? The answer lies not in our algorithms, but in our commitment to transparency. The chain is ready; are we?