Overview
The claim is that Mars Global Surveyor photographed artificial illumination on Mars — lights on the night side, glowing points, evidence of something still running — and that NASA explained it away or declined to release it.
Two things need separating here, because they are usually run together and only one of them has anything behind it.
There is a real and well-documented Mars lights episode, with an image anyone can look at and an explanation that can be checked. It did not come from Mars Global Surveyor. It came from a rover on the ground, in 2014.
And there is a real Mars Global Surveyor story about NASA and Cydonia — one that bears directly on the concealment claim, and that goes the opposite way from the theory.
What Mars Global Surveyor actually did at Cydonia
The Face on Mars came from Viking 1 in 1976: a mesa in the Cydonia region that, at the resolution available and in that particular lighting, looked like a face. Richard Hoagland built a substantial body of work on it, and by the 1990s the standing accusation was that NASA would never point a better camera at it.
NASA pointed a better camera at it, twice, and published both results.
On 5 April 1998 the Mars Orbiter Camera imaged the feature at about 4.3 metres per pixel — roughly ten times the resolution of the best Viking frame. The spacecraft had to be turned off its normal mapping attitude to do it, and it was done because of public interest rather than scientific priority.
On 8 April 2001, at the next opportunity, it was imaged again under cloudless conditions at about 1.56 metres per pixel.
At that resolution it is a mesa: an eroded hill of a kind common in that terrain, whose face-like appearance depends on low sun angle and coarse pixels. The images were released publicly, are still online, and the raw data went into the public archive.
That sequence is the most direct test the concealment theory has ever been given. The prediction was that the agency would avoid the target or suppress the result. It re-pointed the spacecraft, took the picture, and published it — and then did it again three years later at three times the resolution.
The lights that do exist
The famous Mars light image is real, and it is worth being precise about it because the theory usually attaches it to the wrong spacecraft and the wrong decade.
On 2 and 3 April 2014, the Curiosity rover's Mastcam returned frames containing a bright vertical streak near the horizon, which circulated immediately as evidence of artificial light or an underground civilisation.
The explanation is available in the data rather than in an official assurance. Mastcam is a stereo camera with two eyes. The bright spot appears in the right-eye frame and not in the left-eye frame taken less than a second later. Anything actually present in the landscape would appear in both. Something registering in one detector and not the other is in the camera, not on the ground.
The mechanism is routine: a cosmic ray striking the detector deposits energy in the pixels it passes through, and the detector cannot distinguish that from light. Spacecraft cameras collect these constantly, and mission teams treat them as noise. NASA's imaging scientists also offered a glinting rock face as an alternative for some frames, which is the other everyday candidate.
Similar bright spots have appeared in rover imagery repeatedly since, and follow the same pattern.
Why the night-side version does not work
The stronger form of the claim is that Mars was photographed in darkness and bright points therefore had to be emitted rather than reflected light. This runs into how the instruments work.
The Mars Orbiter Camera was built to image a sunlit surface. Its sensitivity and exposure were matched to daylight reflectance off a bright, dusty planet. Pointed at unlit terrain it does not return a dark field with city lights in it; it returns noise. Detecting genuine artificial illumination from orbit is a demanding measurement requiring instruments designed for it, long exposures and careful calibration — the kind of thing done for Earth by dedicated night-vision sensors.
There is also the matter of what the archive would show. MGS returned an enormous volume of imagery over nine years, and it went into public repositories that researchers, students and hobbyists have worked through ever since. A concealed night-side light would have to be absent from a dataset that was never controlled tightly enough to conceal anything.
Why anomalies keep appearing anyway
There is a structural reason Mars produces this material continuously, and it is not credulity.
The planet has been photographed at enormous volume — tens of millions of frames across half a dozen agencies — and almost all of it is public. Every frame contains sensor noise, compression artefacts, dust, shadows and rock. A dataset that large will contain, by simple counting, a great many small features that resemble something familiar, because human vision is built to find faces and shapes in noise and does so reliably even when told not to.
That tendency is well characterised, and it explains the whole genre: the Face, the "spoon", the "doorway", the several rocks resembling animals, and the lights. None of these require anyone to be foolish. They require a large pile of pictures and a visual system that never stops pattern-matching.
The useful test is whether an anomaly survives a second look at higher resolution or from a different instrument. The Face did not. Neither did the rest.
The case the proponents make
The argument is not usually about a specific frame, which is its weakness and its resilience at once.
It is that NASA controls the instruments, the calibration, the processing and the release schedule, so a reader is being asked to accept conclusions they cannot independently verify. Compression artefacts, contrast adjustment and the removal of known noise all involve judgement calls made inside the agency, and the theory reads those as opportunities.
That description of the epistemic situation is fair as far as it goes. What weakens it is that the raw data is in fact published, that non-NASA researchers work with it routinely, and that other nations' orbiters — European, Indian, Chinese, Emirati — have since imaged the same planet with their own instruments and their own institutional interests, including Cydonia. A concealment would now have to be international.
Where it stands
Documented: Mars Global Surveyor operated from 1997 to 2006, imaged Cydonia at high resolution twice, released the results, and resolved the Face as a landform. A bright spot in 2014 Curiosity imagery is explained by the frame in which it does not appear.
Not documented: any Mars Global Surveyor image showing artificial lighting. The claim circulates without a frame number, a release date, or an archive reference — which is unusual for a mission whose entire imaging record is public and searchable.
That absence is the most telling detail. Nearly every other Mars anomaly claim, including the ones that turned out to be mesas and cosmic rays, comes attached to a specific picture you can go and look at. This one does not.