AC and DC both do work in a loudspeaker. The voice coil is a transducer that converts current into displacement.
Applying DC will statically pull or push the cone - you will hear a 'pop' and see the speaker move (that's a handy way to test polarity by the way.) After that there will be no sound. So applying DC to the speaker produces mechanical work (it accelerates some air in a pulse), but just once.
Meanwhile the coil stores magnetic flux and gives off heat. These are also work, just not as useful as sound. (Speaking of which, as @Hearth points out, don't apply a large DC current for a long time as you can damage the driver, especially if it's a plastic cone.)
Applying AC will move the cone back-and-forth. AC power is indeed being converted to mechanical work: moving the coil makes pressure waves in the air. A measure of this electrical to mechanical conversion is 'speaker efficiency', and it is appallingly low - about 0.2 to 2%.
More here: https://engineeryoursound.com/what-does-speaker-efficiency-mean-explained/
Now, if you have an AC signal 'riding on top of' DC, the DC component won't produce sound (except initially), but it will influence the position of the cone. Maybe that's what you mean by DC 'not doing work'.
Now you mentioned 'active' and reactive power. I think you may be slightly confused here. There is real power, which is consumed in the device (as heat, mechanical work etc.) and reactive power, which is the response of a reactive component (L or C) to AC. Reactive power circulates back and forth from source to sink but doesn't do any useful work.
Reactive power only occurs when you have an AC component and reactive elements in your circuit. And that's certainly the case with a speaker: your amplifier can be cranking out lots of watts (volt-amperes) to the speaker but a large portion of that is just circulating back and forth and not being converted to sound.
This is a huge simplification of speaker behavior of course, but this gives the general idea.