Oct 8, 2026 · 20 min
AI could challenge encryption before quantum computers arrive
Are AI Labs Trying to Break Encryption?
The episode frames cryptography as a near-term security concern as AI capabilities advance, with implications for blockchain wallets and broader infrastructure.
- 1Advanced AI systems could eventually threaten cryptographic protocols before practical quantum computers become available.
- 2Ethereum researchers are examining post-quantum security concerns for blockchain wallets and related infrastructure.
- 3Anthropic, Google, and Microsoft are advancing lower-cost models, enterprise agents, and local AI hardware.
Don't miss
The episode’s central warning is that capable AI systems could threaten cryptography before practical quantum computers arrive.
The brief
Brian McCullough opens with a warning that increasingly capable AI systems could eventually break cryptographic protocols, putting blockchain wallets and other security infrastructure at risk.
The concern is notable because it could precede practical quantum computers, shifting the timeline for cryptographic preparedness and making post-quantum security relevant sooner.
Vitalik Buterin, Scott Aaronson, and Justin Drake are tied to the episode’s discussion of Ethereum, quantum computing, computational complexity, and post-quantum cryptography.
The episode then turns to Anthropic’s low-cost Claude Haiku 5.5, Google’s multi-day universal Gemini enterprise agent, and Microsoft’s high-end Surface and local AI developer systems.
Taken together, the stories show AI advancing on two fronts at once: expanding practical deployment while raising harder questions about the security assumptions underneath it.