The short answer: yes, but it is not the most common case.
How it can be? You assume that you reaction has a simple reaction profile with a barrier and a single transition state between reactions and products, plus other approximations. When those approximations fail, the model is not helpful and you have to use a deeper model.
A longer, and hopefully, more nuanced explanation.
For many reactions the dependence of the reaction rate with temperature follows the Arrhnius equation
$$k(T) = A\exp \left(\frac{-E_a}{R\,T}\right)$$
where A, the preexponential factor, and $E_a$ are independent of temperature.
However, you may have that over an extended temperature range many reactions deviate from this behaviour and their behaviour are described better using more complex expressions that behaves as if $A$ and $E_a$ are functions of temperatura. For example, the Kooij equation that has the form
$$k(T) = A T^m \exp\left(\frac{-E_a}{R\,T}\right) $$
As an actual example you can consider the reactión OH + CH$_4$ in the 220-500 K which behaves as Ahhrenius but in the extended interval 220-2000 deviates from that behaviour as the following image shows.

Thus, at present the Arrhenius equation is understood as very useful empirical fitting equation, when it can be applied. But you have to interpret it carefully.
There are some important reactions that have no activation barrier, for example, radical recombinations CH$_{3}\cdot$ + CH$_{3}\cdot$ $\leftarrow$ CH$_3$CH$_3$, or the atmospheric reactions of O($^1$D) -e.g. O($^1$D) + CH$4$-.
Finally, there are some elementary reactions which rate decreases with temperature and fit Arrhenius, thus, its activation energy is negative ($E_a < 0$). For example, the gas phase reaction OH + HNO$_3$ $\leftarrow$ H$_2$O + NO$_3$. Its energy profile does not have a negative barrier, is exhothermic but the reaction profile is complex.


The reaction profile is adapted from the article Brown S.S. et al., J. Phys. Chem. A, 1999 103, 3031-3037