Quantum Computers Are Poised to Crack Encryption. AI Could Beat Them to It.

A facts-only brief, written by a local AI model from this article in The Wall Street Journal, November 6, 2026. It may contain mistakes, and it is not the article: the full story is at wsj.com.

Cryptographers are divided over whether artificial intelligence could undermine the mathematical foundations of modern encryption before quantum computers do. Matthew Green, a professor at Johns Hopkins University, argues that recent AI breakthroughs in mathematics, such as disproving the Exponential Time Hypothesis, suggest that previously assumed “hard” problems may have shortcuts, potentially weakening digital security. He warns that human analysis is no longer the gold standard for assessing cryptographic safety. Conversely, Yehuda Lindell, head of cryptography at Coinbase, contends that speculating about an imminent collapse is irresponsible and that current postquantum algorithms are likely to withstand AI-driven attacks. Lindell believes that even if vulnerabilities emerge, engineers could mitigate them by lengthening cryptographic keys, a change that would slow processing but not render systems unusable.

Background

• Vitalik Buterin, creator of the Ethereum protocol, urged the industry to prepare for “post-AI cryptography” during a conference in Mumbai in early November 2026.
• Green cites specific AI achievements, including a partial solution to the Navier-Stokes equations, as evidence of AI’s growing mathematical capability.
• Lindell prefers that AI labs like OpenAI or Anthropic discover and disclose vulnerabilities, as this allows engineers to patch systems, unlike state actors who might exploit them silently.
• The debate centers on unresolved fundamental mathematical questions regarding whether certain computational problems inherently lack efficient solutions.

Read the full article at wsj.com

More from Saturday, November 7: