Evin McMullen, CEO of Billions, warns that the current crypto architecture is dangerously close to repeating the 'honeypot' mistakes of the Web2 era, but at a much larger scale. While the 2026 market has developed sophisticated tools for asset recovery and chain-halting to mitigate theft, no such mechanism exists for a compromised digital identity. Once a user’s private credentials or biometric data are leaked, that identity is permanently tainted, allowing malicious actors to impersonate the user across the entire decentralized ecosystem.
This vulnerability is particularly acute as the world transitions to an economy driven by billions of autonomous AI agents. These agents act as digital proxies, managing wallets and executing smart contracts on behalf of human users. If the underlying identity architecture remains centralized, these AI agents become high-value targets for hackers. McMullen argues that we are currently handing the keys to a massive, interconnected AI economy over to an infrastructure that was never designed to handle permanent identity security.
From a regulatory perspective, this shift is forcing US agencies like the SEC and FinCEN to re-evaluate what 'consumer protection' looks like in the age of AI. If identity is the ultimate collateral, then a breach is not just a financial loss but a total loss of digital sovereignty. This has sparked a renewed interest in zero-knowledge proofs (ZKPs) and decentralized identity (DID) protocols, which allow agents to verify their authority without revealing the underlying data that could be stolen and misused.
Investors and developers should watch for a pivot toward 'identity-first' security stacks rather than just 'asset-first' firewalls. In the coming months, the focus will likely shift toward projects building permissionless identity layers that can withstand the scrutiny of autonomous agents. The risk of a 'Great Identity Leak' remains a primary bearish concern for the sector, as it could lead to widespread loss of trust in automated DeFi and AI-driven governance models.