Ever since Russia launched its invasion of Ukraine last year, American satellites have proven essential to Ukrainian defense. From SpaceX Starlink satellites keeping communication lines open when ground-based systems were destroyed to spy satellites from Maxar and Planet (PL 0.57%) providing open-source imaging of what invading forces were up to, like it or not, these commercial companies have proven invaluable military assets.

Newsflash: Russia noticed.

Satellite beaming light rays down to Earth.

Image source: Getty Images.

Playing hide-and-seek with satellites

Recognizing that Ukraine's military is using American satellite images to spy on its forces, Russia's naval forces near Crimea have adopted a strategy that is both very old in concept, and rather new in application.

Satellites take pictures from very high up. And the farther away a satellite is from the object it's photographing, the less detail contained in the images. So when the details aren't 100% clear, satellite companies use human analysts -- and increasingly, artificial intelligence -- to tell them precisely what it is they have photographed. But the very fact that satellite images need to be interpreted to understand the information they contain opens up the possibility of misinterpretation, and also of opponents taking steps to fool a satellite.

As NavalNews.com reported last month, Russia is taking such steps, camouflaging several of its warships with a paint scheme that "blacks out" the ship's bow and sterns so as to change its silhouette and make it appear to be a smaller ship in digital images. If successful, such camouflage can trick a satellite (or an attacking drone) into either misidentifying a warship or thinking what it's seeing isn't a warship at all (but rather, say, a harmless fishing trawler). 

Experts note that this kind of camouflage may not be 100% effective -- but also that it doesn't necessarily need to be. AI systems that have been trained to identify ships based on their known size, and that "see" a ship of the "wrong" size, may automatically move on and ignore the target. Human analysts may be better at seeing through the camouflage when given some time to consider it. But even so, that extra time slows down the analysis, and if AI can't be used, this will slow the process even further.

The result: Less timely, less useful intelligence for an opposing military force.

Seeing through the fog of war

Military maneuvers, however, are a game of cat and mouse. One side has a capability, which the other side attempts to counter, prompting the first side to come up with a counter to the counter.

In the context of spy satellites, the counter to Russia's camouflage tactics may be something called "synthetic aperture radar" or SAR -- not a new technology exactly, but not yet common in commercial satellites, either. Rather than taking optical images of targets, as most Earth observations satellites do today, SAR bounces radar beams off of a target to generate an image that can see through clouds (for example), or see through camouflage paint to the solid ship underneath. 

As NavalNews notes, when combined with good artificial intelligence, and with good open source intelligence ("OSINT" -- publicly available information on vessel specifications or last reported locations, for example), images returned by SAR satellites "can help unmask the camouflaged ships" and defeat attempts to fool more run-of-the-mill imaging satellites. For example, NavalNews explains that imagery analysis company Satim (which, according to S&P Global Market Intelligence, is actually a Polish company) has demonstrated how its AI is "unaffected by camouflage" when analyzing SAR images of Russian ships taken by privately held (American) company Umbra Space.

Now, both Satim and Umbra are currently private companies -- i.e. you cannot invest in them. So too are several of the best-known SAR space companies, such as Capella Space and Iceye, for example. But if camouflaging ships and other military assets becomes a trend among hostile militaries (and I see no reason why it wouldn't, if it works), then other companies that are public might well try to duplicate what Satim and Umbra are doing.

Actually, scratch "might." Earlier this year, Maxar Technologies -- a once-public stock recently snapped up in a wave of space stock mergers & acquisitions -- inked a deal with Umbra to secure access to the latter's SAR satellite constellation. Publicly traded space stocks such as Planet and BlackSky (BKSY 0.80%), meanwhile, have both been branching out from ordinary digital photography to introduce hyperspectral imaging into their satellite fleets, which might be another way to defeat Russia's camouflaging efforts. 

At the same time, AI-focused stocks such as Palantir Technologies (PLTR 3.73%) may offer an even better way to invest in this trend, as their tech will be doing the hard work of deciphering digital images no matter how they were generated.

Granted, with a stock price 100 times its annual free cash flow, even Palantir isn't an obviously great investment prospect, while Planet and BlackSky are both entirely unprofitable. Still, stock prices do change over time -- and unprofitable companies can become incredibly profitable when conditions change. If this is the direction in which satellite technology and artificial intelligence technology are moving, investors might want to add all three of these stocks to their watch list.