Well, to disprove such a statement, just give a counterexample.
Alternatively, you can calculate the explicit expression for the number of regions. We do this by induction.
Suppose there are already $(n-1)$ points, and $N_{n-1}$ regions. now we add a point, and gradually draw all the additional $(n-1)$ lines, while tracking the number of regions. Connecting this new point with another one that is $k$ points away (i.e. $k$ points lie between them) creates $k(n-k-2)$ intersections, as the $k$ points on one side and $(n-k-2)$ points on the other side are all connected. These intersections, including the starting/ending point of the segment, divides the new line into $k(n-k-2)+1$ parts. Each part of the segment separates a region. So,
$$N_n = N_{n-1}+ \sum_{k=0}^{n-2}\left(1+ k(n-k-2)\right)=N_{n-1}+n-1+\sum_{k=1}^{n-3}k(n-k-2).$$
We know that $N_2$ is $2$, $N_3=4$, $N_4=8$. Applying this formula gives $N_5=16$, which agrees with my counting. Continue to compute $N_6$ and the result deviates from powers of $2$.
As a side note, the summation above can be shown to be
$$
\frac{n^3}6 - n^2 + \frac{11n}6 - 1,
$$
by realizing it must be a cubic polynomial, and computing the coefficients. And so the expression for $N_n$ reduces to
$$\begin{align}
N_n &= 1+\sum_{m=1}^{n} \frac{m^3}6 - m^2 + \frac{17m}6 - 2\\
&= \frac{n^4}{24} - \frac{n^3}{4} + \frac{23n^2}{24} - \frac{3n}4
\end{align}$$