the question mentions 2 as the rate of change of speed , and we are not supposed to interpret this as magnitude of acceleration. so sqrt(ax^2 +ay^2) is wrong
And also, as prateek said, the expression for d/dt (speed) can be simplified to:
$\mathbf{a}.\mathbf{v} / |\mathbf{v}|$
which makes a lot of intuitive sense too.
this is just the component of $\mathbf{a}$ in the direction of $\mathbf{v}$
The perpendicular component will not be responsible for changing the speed. It will change the direction though.
@Lllt well yes, but how is that relevant here?
@Wolgwang also, as Lllt mentioned, circular motion is a perfect example. In uniform circular motion, the speed is constant. however, there is acceleration of magnitude $v^2/R$