Friction generally spoken causes a change in speed. Physicists call this change — in mathematical terms the speed's derivation — acceleration. While in casual language "to accelerate" means "speeding up", when physicists talk about acceleration they just mean speed change, in any direction.
The underlying reason is that a body's initial speed, including its direction, depends on the observer's viewpoint. The same force therefore can be seen slowing a body down or speeding it up.
Let's take as an example a bullet shot westward along the equatorial line1, with a nozzle velocity of about 465 m/s. The bullet impacts a clay wall which it penetrates for a few centimeters before it gets stuck.
An observer on the grond has no doubt that the friction between clay and bullet slows it down until it reaches standstill. But on the equator the ground is actually moving with about 465 m/s, not quite coincidentally exactly the speed of the bullet: An observer standing "still" above the rotating earth would see the bullet stand still in space until it gets whisked away eastward by the rotating wall.
The important takeaway is that how strong the wall must be to block the bullet, how far the bullet will penetrate, how much it will deform in the process etc. is entirely independent of the reference frame we choose. The same is true for the strength needed for rocket engines to lift a rocket off the ground in Cape Canaveral, the forces on a seatbelt when the car impacts a concrete wall at 65 mph, and all other physical processes. Nothing about them changes when we change the frame of reference, they all are equally valid.2
Bottom line: The same acceleration of the bullet by the wall can be regarded a slowing down or a speeding up or any combination of the two, entirely depending on the observer's reference system. This indicates that the absolute speeds — on which it depends whether friction slows something down or speeds something up — are arbitrary assignments; only the relative speeds matter.
1For the short time we watch the process it is not a big error to ignore that no point on earth's surface is an inertial system (since the earth is rotating), and to ignore that the earth moves around the sun.
2The rotation of the Earth can actually not be neglected when launching rockets (time and distance exceed the small amounts during which it wouldn't matter) but the thrust needed to start depends on it only marginally. Similarly, if we consider significant fractions of lightspeed and significant fractions of the size of the known universe we cannot consider them all equal in cosmological terms: For any given point in spacetime there is a "most natural" frame of reference for which the universe appears the same in all directions, or is isotropic, the comoving frame, but that is beyond the scope of the question.