If a ring of material forms in a plane inclined to the mean plane of a planet's system, forces could act to pull it toward the mean plane, initially causing precession. These forces arise from the quadrupole moment and higher-order terms of the planet's density distribution (usually an equatorial bulge) and the distributions of other satellites and rings in the system. These forces cause precession and do not initially pull the orbital plane into the mean because in order to align fully, it must first lose some orbital energy.
One mechanism for losing orbital energy is through the exchange of orbital energy. For example, the inclinations of nearby more massive objects could be increased slightly.
Another significant mechanism that influences the inclination is inclination dissipation, a type of tidal dissipation. Tidal dissipation, in general, is driven by the tidal bulges on a rotating object. The constant gravitational tugging between the ring particles and the planet's quadrupole moment or nearby moons causes periodic deformations, leading to time-varying forces that can alter the semi-major axis, eccentricity, inclination, and obliquity (tilt). The total losses in orbital dynamics will equal the heating from time-varying deformations, thus conserving energy.
On the other hand, nearly all moons of planets are in the equatorial planes of their host planet's system, even ones that are so far from their host planet that the timescales for inclination dissipation is longer than the age of the solar system. Therefore, these moons could not have been randomly captured at random inclinations and brought down to the mean plane afterwards. They must have formed in situ from dust/rock piles. In the paper "Inclination of satellite orbits about an oblate precessing planet" (1965) by Peter Goldreich, he notes, towards the end,
For example, suppose that in some manner or other all the particles
were put in a common orbital plane which was inclined to the equator
plane by some angle $\alpha$. Then, since the nodes of the individual
particle orbits would move at different rates, the particles would soon
cease to occupy one plane and instead fill a band about the planet.
This band would be symmetrical with respect to the equatorial plane
and would occupy the region outside of a double cone of half-angle
$\frac12 \pi - \alpha$ which has its axis coincident with the plane's
spin axis. Hence, the only planar configuration of particles which
will endure is the one having $\alpha = \frac12 \pi$.
In time, collisions between the particles will turn the band into a thin disk.