(This is an version of reuns answer with added clarity).
Let $\mathcal{A}$ be an associative algebra. Given $w=(A_1,\dots,A_n)\in\mathcal{A}^n$, we define its product $\operatorname{prod}(A_1,\dots,A_n)=A_1\dots A_n$ and given $B\in\mathcal{A}$, we define the multiplicity of $B$ in $w$ as $\operatorname{m}_B (A_1,\dots,A_n)=\#\{j\in\{1,\dots,n\}\mid A_j=B\}$, that is, the number of $B$'s in $w$.
Given $A,B\in\mathcal{A}$, let
$$
p(A,B,n,m) = \sum_{\substack{w\in\{A,B\}^{n+m}\\
\operatorname{m}_A w = n \\
\operatorname{m}_B w = m}}
\operatorname{prod}w
$$
be the sum of the different permutations of $A^nB^m$, so that
$$(A+B)^k = \sum_{m=0}^k p(A,B,k-m,m).$$
Suppose in addition that $\mathcal{A}$ is a normed algebra.
We have, by the multiplicative property of $\Vert\cdot\Vert$,
$$
\begin{align*}
\|p(A,B,k-m,m)\|
&=\left\lVert
\sum_{\substack{w\in\{A,B\}^{k}\\
\operatorname{m}_A w = k-m \\
\operatorname{m}_B w = m}}
\operatorname{prod}w
\right\rVert\\
&\leq
\sum_{\substack{w\in\{A,B\}^{k}\\
\operatorname{m}_A w = k-m \\
\operatorname{m}_B w = m}}
\|\operatorname{prod}w\|\\
&\leq
\sum_{\substack{w\in\{A,B\}^{k}\\
\operatorname{m}_A w = k-m \\
\operatorname{m}_B w = m}}
\|A\|^{k-m}\|B\|^m\\
&=
{k\choose m}\|A\|^{k-m}\|B\|^m;
\end{align*}
$$
so
$$\|(A+B)^k-A^k\| \le \sum_{m=1}^k\|p(A,B,k-m,m)\| \\\le \sum_{m=1}^k{k \choose m} \|B\|^m\|A\|^{k-m} = (\|A\|+\|B\|)^k-\|A\|^k.$$
Now, given an analytic function $f(z) = \sum_{k=0}^\infty c_k z^k$ converging for $|z| < r$ and if $\mathcal{A}$ is a Banach algebra (that is, a normed algebra which is also a Banach space), then the series $f(A)=\sum_{k=0}^\infty c_k A^k$ converges with $A\in\mathcal{A}$, $\|A\|< r$, since it converges absolutely in this case.
Finally, to see the continuity of $f(A)$, if $A,B\in\mathcal{A}$ are such that $\|A\|+\|B\| < r $, then
$$\|f(A+B)-f(A)\| \le \sum_{k=0}^\infty |c_k| \|(A+B)^k-A^k\| \le \sum_{k=0}^\infty |c_k|((\|A\|+\|B\|)^k-\|A\|^k) = g(\|A\|+\|B\|)-g(\|A\|)$$
where $g(z) = \sum_{k=0}^\infty |c_k|z^k$ converges for $|z|<r$.
Hence $f(A+B)\to f(A)$ as $\|B\| \to 0$, that is, we have continuity.