WORLD DEFENSE SHOW 2026 — As the world’s militaries race to take advantage of the breakneck development of artificial intelligence, a Saudi official said today that very soon “there is a decision to be taken” about whether to innovate with “AI-enhanced” defense systems or pursue “AI-native” ones.
Speaking at the World Defense Show outside Riyadh, Majid Algarni, a senior official at the General Authority for Defense Development, explained that AI-enhanced systems are available today: current tech that can integrate AI in a relatively limited capacity.
“We can see that in the whole show now,” he said.
But he said a day is coming shortly when Saudi Arabia would have to consider the benefits and drawbacks of “AI-native” platforms that are built from the ground up with AI baked in. Or, as he put it, “the full value chain, from chips to data to models and then to agentic stuff.”
“This is a big leap, requires a lot of readiness and a big investment. But this is going to be a key changer, a force multiplyer that we need to be taking into consideration,” he said.
It must be carefully considered, however, because he and fellow panelists acknowledged the risks and fears that naturally come with the idea of ceding too much military decision-making to AI. Lawrence Schuette, director of research and technology programs at American defense giant Lockheed Martin, said it comes down to a question of trust.
“Go back and ask ChatGPT how many ‘r’s there are in the word strawberry. It’ll tell you two. But your kindergartner will tell you three, and your kindergartener is correct,” he said during the panel discussion. “So what other hallucinations are occuring in your AI system that’s making a decision for you?
“So the first rule is, it can’t make the decisions. It has to be a piece of that decision-making process, and then as you learn to trust it, and it trust you, you’re going to be able to make faster decisions at machine speed but with human accuracy,” he said. (In a quick Breaking Defense test, a ChatGPT model answered the strawberry question correctly, but other hallucinations have been well documented.)
Saudi Arabia is hardly the first country to weigh the benefits and risks of more AI integration. An exercise in September by the US Air Force showed that AI agents wrote up battle plans some 400 times faster than humans could, though some were infeasible. The Pentagon has also deployed versions of AI chatbots military-wide.
“Becoming an ‘AI-First’ warfighting force requires more than integrating Al into existing workflows,” US Secretary of Defense Pete Hegseth wrote in an “Artificial Intelligence Strategy” memo [PDF] last month. “It requires re-imagining how existing workflows, processes, TTPs [techniques, tactics and procedures], and operational concepts would be designed if current AI technology existed when they were created — and then re-inventing them accordingly.”
US military planners have long said that a human will always be “in the loop” to guard against mistakes when it comes to potentially lethal operations, and Algarni echoed the point today.
“Humans need to be in the loop to make sure that [for] every single critical killing chain action, a human is there to ensure full control,” he said.
While AI is the “first priority” for tech under Riyadh’s broader Vision 2030 development initiative, Algarni said, “We do care about responsible AI in the military.” He noted that Saudi Arabia has signed onto international AI safeguard agreements.
Still, by the end of the panel, panelists seemed resigned to the idea that in the not-too-far future there won’t be much of a choice between enhanced AI and native AI — native AI will be the norm.
“For the long term, the human-machine interface will be taken to another league,” Algarni said.
PHOTOS: World Defense Show 2026
Gallery
