Haseeb Qureshi Says Crypto Was Built for AI, Not Humans

Haseeb Qureshi, managing partner at Dragonfly, argues that crypto’s persistent friction is not the result of poor engineering but of a deeper mismatch between blockchain architecture and human behavior. In a detailed post on X, he suggested that many of crypto’s perceived failure points stem from the simple fact that humans were never the ideal primary users of the system.

According to Qureshi, the early vision of crypto imagined a world where smart contracts would replace legal agreements and courts, with property rights enforced directly on-chain. Yet that transformation has not occurred. Even crypto-native investment firms still rely on traditional legal frameworks when closing deals. Dragonfly, for example, signs conventional legal contracts alongside any on-chain vesting agreements, reflecting a continued dependence on established legal systems.

A Structural Misalignment Between Humans and Code

Qureshi believes the issue is social rather than technical. Blockchain systems function deterministically, enforcing rules exactly as written. Humans, however, are inconsistent, emotional, and prone to error. Traditional banking systems evolved over centuries to accommodate human mistakes, fraud, and misuse. Crypto, by contrast, was architected around rigid execution and automated enforcement.

Features such as long cryptographic addresses, blind transaction signing, immutable transfers, and automated contract execution demand constant vigilance. Users are expected to verify domains, audit contract logic, and avoid spoofed addresses every time they interact with the system. In reality, human behavior does not align with that expectation. Mistakes happen, and in crypto, errors are often irreversible.

This disconnect, Qureshi argues, explains why crypto still feels intimidating to many users in 2026. What appears to be broken design may instead be evidence that the system was never optimized for human intuition in the first place.

AI Agents as Crypto’s Natural Users

Qureshi suggests that artificial intelligence agents may be better suited to crypto’s deterministic framework. Unlike humans, AI systems do not experience fatigue or skip verification steps. They can analyze smart contract logic, simulate edge cases, and execute transactions without hesitation. In this sense, crypto’s rigid, rule-based structure may be perfectly aligned with machine logic.

He envisions a future built around “self-driving wallets,” where AI agents manage financial activities on behalf of users. In such a model, autonomous systems could transact directly with each other on always-on, permissionless blockchain networks. Rather than struggling with complexity, AI agents would treat crypto’s strict execution rules as a well-defined specification.

AiFi and the Next Phase of Adoption

The idea that crypto is better suited for machines than humans is gaining traction among industry voices. Ryan Adams, founder of Bankless, recently argued that what is often labeled as poor user experience for humans may actually be optimal UX for AI agents. He predicts that billions — and eventually trillions — of AI-driven wallets could emerge, pushing crypto markets beyond $10 trillion in value. This narrative, sometimes referred to as “AiFi,” draws comparisons to the early days of decentralized finance before its rapid expansion.

Similarly, Changpeng Zhao has stated that crypto is likely to become the native currency layer for AI agents, reinforcing the idea that machine-to-machine economies may rely on blockchain infrastructure.

Structural Limits Remain

Despite the strength of the machine-native thesis, significant constraints persist. Liability for AI-driven financial actions still ultimately rests with humans or institutions, keeping traditional legal systems relevant. Deterministic smart contracts may reduce ambiguity, but they do not eliminate exploits, governance failures, or systemic risks.

Moreover, if AI becomes the dominant interface layer, crypto could increasingly function as backend infrastructure rather than as a parallel financial system visible to everyday users. In that scenario, blockchain technology may become foundational — but less directly interacted with by humans.

Whether crypto evolves into the operating system for autonomous agents or remains a niche financial layer will depend on how AI integration unfolds. For now, Qureshi’s argument reframes the debate: perhaps crypto’s friction is not a flaw, but a signal that the true users of the system have yet to fully arrive.