Under attack from the militant Houthi group in neighbouring Yemen, Saudi Arabia has asked numerous allies for help.
That includes the United States, even though media reports say US President Donald Trump refused when Saudi ruler Mohammed bin Salman asked him to take direct military action against the Houthis.
However, the US is providing Saudi Arabia with intelligence, including a targeting tool that uses artificial intelligence (AI). A September report by US broadcaster CNN, quoting several unnamed American officials, said that between 100 and 200 US soldiers were in Saudi Arabia as part of a new joint force command to facilitate this.
The US personnel were helping the Saudis "use a targeting tool that allows military officers to see in real time what is happening on the battlefield", CNN reported.
Officials said the system in Saudi Arabia was similar to the Maven Smart System (MSS), which the US military has also used to identify targets in Iran, Iraq and Syria.
What is Maven?
MSS emerged from Project Maven, the US military's flagship artificial intelligence programme, first launched in 2017. MSS itself is an AI-enabled digital platform that "aggregates, organises and visualises vast streams of intelligence, surveillance and reconnaissance data", analysts at the Center for Strategic and International Studies in Washington explained in a June article.
The software system identifies patterns in data from sources such as live satellite feeds and drone cameras, then flags potential targets, including possible rocket launchers, tanks and suspicious vehicle convoys.
Previously, human military analysts would have had to examine a wide range of different sources to identify such targets, the CSIS researchers wrote. This was difficult and, in the case of hours of satellite or drone footage, almost impossible. MSS now brings hundreds of sources together, analyses them and flags possible targets. It also calculates the best method of striking a target.
Tech commentator Bilawal Sidhu, a former Google product manager, has described the MSS set-up as being like a "Google Earth for war".
Using MSS dramatically speeds up what is known as the "kill chain" — the time it takes to detect a target, determine whether it poses a threat and order a strike.
When the US military used MSS during its attack on Iran in late February, the AI-powered system facilitated strikes on more than 1,000 targets within the first 24 hours of the war. Previously, American soldiers had been striking about 100 targets a day.
A faster kill chain?
Until recently, the long-running civil war in Yemen had been comparatively calm following an informal ceasefire negotiated in 2022.
The Houthi rebel group has controlled large parts of northern and western Yemen since 2015. Yemen's internationally recognised government and other groups control territory across the south and east. The latter is supported by Saudi Arabia and the United Arab Emirates.
Previously, human military analysts would have had to look at a wide array of different sources to try to find those, the CSIS researchers wrote. It was difficult and, in the case of hours of satellite or drone footage, almost impossible. But MSS now brings hundreds of sources together, analyses them all, then flags possible targets. It also calculates the best method of hitting the target.
The Houthi group's latest military campaign, which began in September, has seen it take further control of Yemen's coast. It has also stepped up attacks on Saudi Arabia, striking the country's main airport and oil facilities. So far, however, military campaigns against the Houthis by Saudi Arabia, the United States and others have all failed to significantly weaken the group.
This is partly because the Houthis take a "zero-sum approach", Farea al-Muslimi, a research fellow specialising in Yemen at the UK think tank Chatham House, recently told DW. The group's founder argued that contemporary Muslim societies were too "soft" and that poverty and hardship were divine tests, al-Muslimi said.
"They don't care if the Israelis bomb the airport or the US bombs a hospital," the expert said. "It actually serves them very well because their ideology is all about martyrdom and sacrifice."
MSS's major issues
There are technical issues with MSS that users have noted in the past and that could also prove problematic in Yemen.
The Houthis tend to be highly mobile, using truck-mounted ballistic missiles and small drone swarms while operating from hills and canyons in rugged countryside. Fighters might have moved on before they can be attacked, even with the help of AI-assisted targeting.
Although MSS is known to work best in relatively unpopulated areas, factors such as changes in desert terrain, dust storms, high winds, snowfall and dense foliage — as well as deliberately planted decoys — can reduce its success rate.
There is also a worrying ethical aspect to shortening the kill chain. A human is always supposed to verify a target identified by an AI-assisted tool and decide how to respond.
But it does not always work that way. For example, among the sites targeted by MSS in Iran on the first day of the war was a school. A US missile killed almost 200 people at a primary school in southern Iran, most of them children aged between seven and 12.
Could AI defeat the Houthis?
During its military campaign in Gaza, Israel — arguably the Middle Eastern country with the most advanced military AI capabilities — also used AI-assisted targeting. An Israeli system known as the Gospel, or Habsora, is the closest equivalent to MSS in how it works.
However, as documented elsewhere, including by Israeli and British journalists and in the recent documentary Naza, the use of such AI targeting tools has also proved highly problematic because of the civilian casualties that can result.
"On the one hand [in Gaza], AI has enhanced commanders' situational awareness, lessened the human costs of tunnel mapping, improved the speed of military strikes, and boosted the survivability of troops," analysts Frederic Wehrey and Andrew Bonney wrote for War on the Rocks, a specialist media outlet. "Yet the benefits afforded by such functions have been overshadowed by mounting civilian deaths reportedly resulting from this technology's safeguards being loosened or eliminated."
For these reasons, Antony Loewenstein, an independent journalist based in Australia who has researched the impact of Israel's AI-assisted technology, does not believe similar technology will lead to the defeat of the Houthi group.
"I don't think there's much evidence at all that using an AI tool is going to make much difference on the ground, except for causing huge amounts of civilian casualties," Loewenstein, author of The Palestine Laboratory: How Israel Exports the Technology of Occupation Around the World, told DW. "And I think what the US and the Saudis will find is that the Houthis have an incredibly high tolerance level for destruction."
Edited by: Carla Bleiker
Cathrin Schaer is a freelance journalist based in Berlin, Cathrin Schaer's work has been published in a variety of media, including the New York Times, The Atlantic, Al Jazeera and The Guardian, among others.
Disclaimer: This article first appeared on DW, and is published by special syndication arrangement.