Gravitational attraction and electrostatic attraction/repulsion are intrinsic properties of matter, any particle (electron, proton) for some unknown reason can produce KE at a distance.
This is an observational fact.
But magnetic attraction/force is not an intrinsic property of matter, a charged particle generates a magnetic field/flux and a magnetic force only when it is moving: higher velocity = much higher force.
This is one of the two ways where magnetic force has been detected :
Fundamentally, contributions to any system's magnetic moment may come from sources of two kinds: (1) motion of electric charges, such as electric currents, and (2) the intrinsic magnetism of elementary particles, such as the electron.
Thus the statement does not hold. Magnetic forces are as intrinsic as electric, just harder to model.
The definition of KE says that it is 'work done to accelerate an object', energy spent only to make it move, and does not mention the generation of other forces/energy or doing 'extramural' work.
Work-Energy Principle
$$W_\text{net} = \frac{1}{2} mv_\text{final}^2 -\frac{1}{2} mv_\text{initial}^2 $$
The change in the kinetic energy of an object is equal to the net work done on the object.
That is the statement. It does not say a lot of things that are irrelevant, to the simple conservation of energy form given above. How this is accomplished is a matter of study
Let us take your last proposition:
An electron is travelling at high speed (say,0.9 C). If is moving near another electron, (proton, positron or a live wire) it can make anything move, acquire KE, it can do work.
One cannot be changing the initial conditions randomly. Each change , for example acquire more kinetic energy or lose energy to work is a different boundary condition problem.
When v approaches c, the attractive magnetic force gets so great that it equals the huge electrostatic repulsion.
The simple problem is that the electron travels with energy v. If it is being accelerated it is a different boundary condition problem to be solved. Let us look at the simpler boundary condition problem.
Let us take two electrons traveling in parallel in vacuum at constant velocity. Their individual kinetic energy to start with is constant, no work is done . If they are close enough to feel the electrostatic potential of each other , the kinetic energy can change, and the energy is supplied by the potential energy given when the tracks were designed/forced to be close to each other.
If they are close enough and move fast enough to feel the magnetic force of the motion also
$$\vec{F}= \underset{\text{Electric force}}{q\color{red}{\vec{E}}} + \underset{\text{Magnetic force}}{q\vec{v}x\color{blue}{\vec{B}}}$$
again the energy supplied for any change in the motion/kinetic_energy of the electrons is due to the initial conditions that generated the trajectories. Those initial conditions were imposed using energy which is given up when the potential energy of the two particle system changes.
Both electric fields and magnetic fields store energy. For the electric field the energy density is
$$\eta_E = \frac{\text{energy}}{\text{volume}} = \frac{1}{2} \epsilon E^2$$
This energy density can be used to calculate the energy stored in a capacitor.
For the magnetic field the energy density is
$$\eta_B =\frac{\text{energy}}{\text{volume}} = \frac{1}{2} \frac{B^2}{\mu} $$
which is used to calculate the energy stored in an inductor.
In the simple two electron in parallel fast trajectories example, the electric field energy density is given by the charges of the two electrons and the constraint of the trajectory, and the magnetic field in addition has the initial velocity as input constraint. In interaction the potential energy of the initial two particle system is transformed to the changes in the kinetic energy of the final two particle system . The potential energy was given by the initial setup of the two particle system.
To keep electrons in a steady v velocity beam in an accelerator, energy is continuously supplied by the beam line magnets etc. One does not get the beam up to a velocity and then let it fly at steady v, because it will disperse due to the forces shown above, for this reason focusing with magnets is used at the last part of the beam as seen in the linked plot.
Energy is conserved.