This is a guess. Finely divided iron is pyrophoric, the intense flashes in fireworks are due to metal powders burning in air (magnesium and I think aluminum as well). Aluminum should be more pyrophoric than iron when chipped into small pieces. It is however covered with a layer of aluminum oxide (Al2O3) otherwise known as corundum which is extremely hard and chemically inert (ruby is corundum with a small amount of chromium added). Commercial planes are made of aluminum and to give the surface extra protection the oxide skin is made thicker by electrochemical methods. So my guess is if you can scratch below the oxide layer you may be able to get sparks of Al reacting with air. The oxidation reaction may be so quick however that the oxide layer is immediately formed when fresh aluminum is exposed to air. It will not spark if the oxide layer remains intact.
Aluminum is often referred to as a "non-sparking" material, and this property is frequently listed for aluminum in the Periodic Table of Elements. The principal interest in using non-sparking materials is to avoid ignition of combustible or explosive materials from an "impact generated spark".
One will often see advertisements for "non-sparking tools" made from aluminum alloys. Does this mean these "non-sparking tools" are safe in hazardous environments, or that they will not spark? Not necessarily - ignition by a chemically generated spark may also occur - depending on what the aluminum strikes - i.e. rusty steel (ferrous oxide).
What about grinding steel versus grinding aluminum? We have often seen the showers of sparks given off when steel is pressed against a grinding wheel. Does this mean it is safe to grind aluminum? Again not necessarily. Apart from the dangerous build-up of friction (heat) that can occur - you can generate a very dangerous mixture of iron and aluminum dust called THERMITE.
Here's an industrial datasheet on aluminum.
Let's not forget that aluminum has another important property as an electrical conductor. In other words it is possible for aluminum to conduct and discharge static electricity, again creating a spark-type ignition source. Up until now we have discussed solid aluminum. If aluminum is processed into a fine powder it oxidizes, and there are hazards associated with the handling of this material.
In conclusion - you would probably freeze to death before you could start a campfire by striking two pieces of aluminum together. Steel and flint would be a far more successful combination. On the other hand I wouldn't bet my life that aluminum is a completely non-sparking material.
I am associated with production of solid fuels, where I came across the safety requirements of Tools specified by OSHA, CCOHS and other safety controlling organisations. Everywhere, it is mentioned that non-sparking tools, specifically made of Copper-beryllium alloy or titanium are to be used for safe machining and end trimming but after going through manuals of very non-sparking tools suppliers, it became clear that there is nothing that is non-sparking. even High Speed Steel (HSS) is claimed to be non-sparking as compared to tool steel. All these so called non-sparking tools are basically spark-reducing tools and there use by no means can ensure absence of impact sparks. Of course the peculiar characteristics observed in so called spark reducing materials, are their low tensile strength and high thermal conductivity. All these non-sparking materials deform before attainment of a temperature needed for the generation of spark. This is only my suggestion, observation and practical experience. If Aluminium suits in this specification, we may mark it also as "NON-SPARKING".
The term "non-sparking" or "spark reducing" tools has to do with copper beryllium or copper aluminum alloys or any of a host of others being cold spark materials. The standard compares the sparks generated by friction or impact (which they all make to some degree) to their ability to ignite carbon disulfide (CS2) which has the lowest known auto-ignition temperature (290 F if I remember correctly). Finding a technical document that discussed the spark temperature of aluminum and correlating that with the reference to the CS2 (likely found in the OSHA regs, but it's been a long time since I looked it up specifically so it might take some digging), might give you the information you need to prove reasonably enough that aluminum is considered non-sparking.