According to Huehnergard, Akkadian had a phonemic glottal stop. This makes sense, given the language's heritage.
However, he doesn't seem to mention it anywhere in the chapters on orthography, and I certainly have never seen cuneiform glyphs for (e.g.) ʔa as opposed to a (*).
So, how did Akkadian scribes mark glottal stops? And if they didn't—that is, if they wrote naʔdum "attentive" as na-du-um with no indication of the ʔ—how do we know the glottal stop existed at all, and didn't just disappear in Akkadian like Proto-Semitic *ʕ and *h did?
(*) Kloekhorst argues that certain glyphs could indicate the presence of a glottal stop in Hittite, the reflex of PIE *h₁ in certain environments, but this doesn't seem to be a mainstream view.