There is clearly plenty of evidence that shows that the magnetic properties are far more radiation-resistant than biological molecules, and can persist at doses that would destroy DNA. The term "unaffected" is rather absolute, but I interpret the claim to mean that the quality of science done through magnetic analysis of the samples, after they have been irradiated, would be unaffected. Theoretically there could be an analytical technique that is especially sensitive to the sample's magnetic properties, and such a technique might be affected, but I am not aware of any such technique. Even if it were affected, it might be possible to control for the effect and still tease out useful findings. I do not think that irradiation would completely erase all magnetically encoded information. Instead, it would just degrade the signal-to-noise ratio slightly. Additionally, the sample will likely have been naturally irradiated for millions of years prior to retrieval, so we're talking about here is just an incremental amount of additional radiation for sample sterilization. Therefore, I believe that for all practical purposes, this claim is well supported.

That said, I do think that this claim would benefit from an additional review by someone with expertise in paleomagnetism or magnetic materials science.