I don't have the code for tex mode with me now, but yes, it is possible, in fact all too common, IMHO, that the same font-lock face is reused for something quite different. If that is the case here, i.e., if the tex-mode code hard-codes the face it uses as the same face, then you are out of luck (modulo rewriting that code to use a different face).
To see what face is actually used, in your tex-mode buffer, put the cursor on the text in question and hit
C-u C-x =. That will display lots of info about the position and character, and near the bottom it will tell you about any text properties, including property
face, that are on that character. If it says
font-lock-type-face then yes, the same face is used.
If that is not the case, so you see a different face name, then the problem is perhaps only that that face's definition inherits from
font-lock-type-face. In that case, you're in luck: just customize that face, whatever it is.
In the out-of-luck case, consider also using
M-x report-emacs-bug to suggest that Emacs Dev not simply reuse the same face,
font-lock-type-face, but instead they create a tex mode-specific face for this. They can have that face inherit from
font-lock-type-face if they want, but they should not just reuse that face. This is my opinion, not necessarily shared by Emacs Dev -- just letting you know.