AI could help researchers build quantum computers capable of threatening public-key encryption while helping organizations find vulnerable cryptography, according to an analysis by Troy Coienth Troublefield reported Oct. 21 by The Quantum Insider. The paper brings together existing research and security assessments rather than presenting new experimental findings.
A sufficiently capable quantum computer running Shor’s algorithm could undermine affected public-key systems, including RSA and elliptic-curve cryptography. Today’s processors remain too small and error-prone to carry out those attacks against widely used cryptography, the report says.
Why prepare before an attack is possible?
The analysis warns that attackers could save encrypted communications now and read them with a sufficiently capable future quantum computer. Information that must remain private for years could therefore need protection before such a computer exists. The paper notes that publicly available evidence does not establish who is collecting encrypted material or how much has been collected.
AI on both sides of the problem
On the threat side, Troublefield suggests AI could improve quantum error correction. Machine-learning decoders might identify errors more accurately or quickly, supporting efforts to make quantum calculations more reliable. That is a possible contribution to developing capable machines, not a demonstrated attack.
On defense, the analysis identifies automated discovery as a potentially practical near-term use of AI. Such tools could help security teams find vulnerable cryptography across applications, certificates and connected devices, then organize its replacement. The Quantum Insider reports that the National Institute of Standards and Technology finalized its first three post-quantum cryptography standards in August 2024. Troublefield recommends prioritizing long-lived sensitive information, keeping a cryptographic inventory, protecting AI infrastructure and designing systems so cryptographic methods can be replaced.
Discovery tools are not a complete answer: the paper says independent public evaluations of commercial products’ accuracy and coverage remain limited. Its proposals offer a , not new experimental findings or a demonstrated attack on widely used cryptography.
