Securing generative artificial intelligence with parallel magnetic tunnel junction true randomness
This addresses security vulnerabilities in generative AI systems for applications requiring high randomness, though it is incremental as it builds on existing spintronic technology.
The paper tackled the problem of predictable patterns in generative AI models due to deterministic pseudo random number generators, by embedding hardware-generated true random bits from spin-transfer torque magnetic tunnel junctions, resulting in up to 18.6 times reduction in insecure outputs for a GAN on CIFAR-10 compared to a baseline.
Deterministic pseudo random number generators (PRNGs) used in generative artificial intelligence (GAI) models produce predictable patterns vulnerable to exploitation by attackers. Conventional defences against the vulnerabilities often come with significant energy and latency overhead. Here, we embed hardware-generated true random bits from spin-transfer torque magnetic tunnel junctions (STT-MTJs) to address the challenges. A highly parallel, FPGA-assisted prototype computing system delivers megabit-per-second true random numbers, passing NIST randomness tests after in-situ operations with minimal overhead. Integrating the hardware random bits into a generative adversarial network (GAN) trained on CIFAR-10 reduces insecure outputs by up to 18.6 times compared to the low-quality random number generators (RNG) baseline. With nanosecond switching speed, high energy efficiency, and established scalability, our STT-MTJ-based system holds the potential to scale beyond 106 parallel cells, achieving gigabit-per-second throughput suitable for large language model sampling. This advancement highlights spintronic RNGs as practical security components for next-generation GAI systems.