Shayan Sardarizadeh
BBC Verify senior journalist

As we’ve previously reported, satellite imagery is an important tool in our work at BBC Verify as it gives us visual evidence from locations where access is either impossible or restricted.

It also provides a birds-eye view of an area of interest, for instance helping us monitor the extent of damage in a war zone over time.

Earlier this week, we used high-resolution satellite images of Russian airbases to analyse damage inflicted on multiple aircraft by Ukrainian drones.

While the footage we’ve verified shows real damage, in recent months (and with the rapid advancement in AI) we’ve also seen the rise of a new phenomenon as a tool of misinformation: AI-generated satellite imagery.

Here’s a very recent example shared online claiming to be satellite imagery documenting damaged Russian aircraft.

But this is fake, created using AI techniques.

We can establish this by zooming in to look at the detailed structure of individual aircraft. Multiple physical anomalies are visible, including asymmetrical wings, misaligned engines and other features that do not make physical sense.

BBC Verify has observed similar AI-generated and misleading satellite images shared in relation to news events from North Korea, Iran, and the Israel-Gaza war.

While authenticating generative AI content is increasingly becoming difficult, zooming in to spot inconsistencies and verifying sources are good places to start.