Adjoint Representation
Definition. $G$ acting on its own Lie Algebra: $\mathrm{Ad}_g(X) = gXg^{-1}$ (matrix groups). Differentiating gives the algebra acting on itself: $\mathrm{ad}_X(Y) = [X, Y]$.
Working examples. $U(1)$ is abelian ⇒ $\mathrm{Ad}$ is trivial ($gXg^{-1} = X$). $SU(2)$: $\mathrm{Ad}$ is the spin-1 (3-dimensional) representation — and gives the double cover $SU(2) \to SO(3)$.
Why it matters here. Under a Gauge Transformation, the Curvature 2-Form transforms in the adjoint: $F \mapsto g F g^{-1}$. For $U(1)$ this means $F \mapsto F$ — the electromagnetic field strength is gauge-invariant, which is why Landau can treat $F_{ik}$ as directly physical. In Yang–Mills, $F$ is only ad-covariant, not invariant; you must trace over it ($\mathrm{tr}\,F\wedge\star F$) to build observables. The abelian simplicity of EM is the triviality of $\mathrm{Ad}$.
Also: gauge fields themselves "transform in the adjoint" (plus the inhomogeneous $g^{-1}dg$ term) — gluons carry color charge, photons carry no electric charge. Same fact.
Related: Lie Algebra · Group Representation · Curvature 2-Form · Yang-Mills Theory · Module 10 - Representations and Matter Fields
Sources: Woit (adjoint rep chapter); Sternberg Ch. V.