Violation of causality in f(T) gravity

[Abridged] In its standard formulation, the $f(T)$ field equations are not invariant under local Lorentz transformations, and thus the theory does not inherit the causal structure of special relativity. A locally Lorentz covariant $f(T)$ gravity theory has been devised recently, and this local causality problem has been overcome. The nonlocal question, however, is left open. If gravitation is to be described by this covariant $f(T)$ gravity theory there are a number of issues that ought to be examined in its context, including the question as to whether its field equations allow homogeneous G\"odel-type solutions, which necessarily leads to violation of causality on nonlocal scale. Here, to look into the potentialities and difficulties of the covariant $f(T)$ theories, we examine whether they admit G\"odel-type solutions. We take a combination of a perfect fluid with electromagnetic plus a scalar field as source, and determine a general G\"odel-type solution, which contains special solutions in which the essential parameter of G\"odel-type geometries, $m^2$, defines any class of homogeneous G\"odel-type geometries. We extended to the context of covariant $f(T)$ gravity a theorem, which ensures that any perfect-fluid homogeneous G\"odel-type solution defines the same set of G\"odel tetrads $h_A^{~\mu}$ up to a Lorentz transformation. We also shown that the single massless scalar field generates G\"odel-type solution with no closed timelike curves. Even though the covariant $f(T)$ gravity restores Lorentz covariance of the field equations and the local validity of the causality principle, the bare existence of the G\"odel-type solutions makes apparent that the covariant formulation of $f(T)$ gravity does not preclude non-local violation of causality in the form of closed timelike curves.