Possibly the journalist mixed two totally separate stories, hydrogen embrittlement and corrosion fatigue. That would fit the symptoms and the journalists visual observations of testing.
I suppose H2 damaged steel, could lead to cracks, that when put underwater under tension would cause faster than expected corrosion fatigue like a couple years instead of a couple centuries, they're not totally unrelated.
(edited to rephrase corrosion doesn't cause brittleness but brittleness can speed corrosion in the above scheme)