Quantum computing offers incredible potential, but due to its susceptibility to noise, it has some practical limitations. Now, Rice University computer scientists are tackling “quantum adversaries” who seek to exploit these weaknesses purposely.
The Rice team’s work has developed algorithms targeting the noise that seeps into quantum data, whether it is inherent or intentional. This could be a significant step toward enhancing the reliability of quantum computing and bringing the technology into practical use.
Quantum Computing
While traditional computing stores information in bits, the 1s and 0s that make up binary code, quantum computing operates on qubits. These pieces of data exist in a quantum state, where multiple probabilities coexist, instead of a single value. Measuring the qubits’ quantum state reduces it to a single, random outcome.
“According to the laws of quantum mechanics, observing a quantum state often ‘destroys’ it, resulting in a random measurement that only reveals partial information about that state,” said lead author Yuhan Liu. “Quantum state learning studies how to accurately translate quantum information by using multiple copies of the quantum state.”
Quantum learning is an essential tool for investigating quantum computing. The technique reconstructs a quantum state from multiple instances of the same state. This enables researchers to validate algorithms and clock hardware in the pursuit of reliable quantum computing.
Noisy Interference
That reliability is the central issue plaguing more widespread adoption of the technology, which remains in a development phase called “noisy intermediate-scale quantum” or NISQ. Even minor environmental disturbances can impact quantum technologies, leading to measurement errors and corruption. Mitigating noise in the data is the next step for quantum computer researchers.
Quantum computer specialists have developed various ideas for modeling noise, either uniformly or randomly. The Rice team’s innovation stems from including targeted interference from quantum adversaries to produce the most realistic framework yet.
“Our model is strong in the sense that it also considers nonphysical and potentially malicious factors that may affect the system,” co-author Nai-Hui Chia said. “Our goal here was to see if we could design a good algorithm to certify devices or do other tasks such that we would be secure against a deliberate attack by an adversary.”
Quantum Results
The framework delivered mixed results when it came to determining the technology’s vulnerability to a quantum adversary.
“The bad news is that for some states, learning under adversarial noise is nearly impossible,” Liu said. “An adversary only needs to change an exponentially small fraction of the states or measurements to totally fool any learning algorithm.”
“The good news is that for a large class of well-structured states that are frequently used in many quantum algorithms, it is possible to achieve reasonably good accuracy, even when the noise is added maliciously,” Liu added.
Pushing Beyond The Noise
This recent work is part of a broader effort to advance quantum computing technology beyond the NISQ stage, which will involve developing more robust noise mitigation strategies. The Rice team suggests that, in addition to algorithmic solutions, fine-tuning the physical systems will also be necessary. Material science is producing interesting solutions, such as new two-dimensional materials with properties that decrease noise and increase stability.
Still, the software that these advanced computing systems operate on will be critical to avoiding quantum adversaries and random background disturbances. The Rice team’s framework provides a new tool for tuning algorithms to safeguard the delicate quantum states required for the stable operation of quantum computing technologies.
The paper, “Adversarially Robust Quantum State Learning and Testing,” will be delivered at the 2025 IEEE Symposium on Foundations of Computer Science in December 2025.
Ryan Whalen covers science and technology for The Debrief. He holds an MA in History and a Master of Library and Information Science with a certificate in Data Science. He can be contacted at ryan@thedebrief.org, and follow him on Twitter @mdntwvlf.