It’s been a big couple days in the increasingly active defense tech sector.
Today, Microsoft and Palantir announced they will bundle their AI services for U.S. military and intelligence, with Microsoft contributing GPT-4 and other OpenAI models, and Palantir throwing in its targeting, data-analysis, and action-proposing tools. This is the first partnership around deploying Microsoft Azure OpenAI Service in classified environments, Palantir CTO Shyam Sankar noted in a statement—though Microsoft, on its own, did deploy GPT-4 to its Azure Government Top Secret environment in May. And of course, the relevant government agencies still need to bite.
Microsoft and Palantir say their combined services could be used in logistics, contracting, and action planning, among other things.
As Military.com reported last month, there is little agreement among the armed forces regarding the use of generative AI. Some have security concerns, such as the Space Force (which has paused the use of the technology) and the Navy. The Army is reportedly more open to generative AI, though it still lacks clear rules to govern its use.
Earlier this week, Palantir also delivered a Tactical Intelligence Targeting Access Node (TITAN) prototype to a military base in Washington State, Axios reported. TITAN is a wheeled ground station system that uses AI to assist targeting decisions, based on data collected from sensors in the air, on land, and in space.
Palantir won the $178.4 million Army deal a couple of months back and will develop 10 TITAN prototypes for testing. Bearing in mind that we’re talking about what’s generally seen as a data company, it’s remarkable that it’s now delivering hardware—thanks to subcontractors such as Northrop Grumman and Anduril.
Speaking of Anduril, Palmer Luckey’s defense tech startup just pulled in a fresh $1.5 billion in funding, as it takes aim at the more traditional likes of Northrop Grumman and Lockheed Martin—Anduril recently won out against these rivals for an Air Force contract to develop and test prototypes of small, unmanned fighter jets.
The new round, co-led by Founders Fund and Sands Capital, values Anduril at $14 billion, which TechCrunch points out is a darn sight more than the $8.5 billion valuation the firm boasted less than two years ago. It’s also more than the $12.5 billion valuation that Anduril was reportedly seeking less than three months ago.
As my colleague Luisa Beltran wrote a couple weeks ago, defense tech—particularly when it involves hardware—is becoming a more attractive bet for investors than used to be the case. That’s partly because software is looking less lucrative in a scenario where AI becomes a capable coder, and partly because of the wars in Ukraine and Israel/Palestine, plus geopolitical tensions between the U.S. and China.
See also: Germany’s Helsing raising $489 million in July, at a valuation of $5.4 billion. The defense AI firm’s previous round in September last year pulled in less than half that amount, at less than a third of the valuation.
There are of course serious ethical questions about the use of AI for target selection, and not much in the way of an international legal framework for controlling the phenomenon. Reported deployments in Gaza and Ukraine make these questions more urgent than anyone might have expected a couple years back—and so do deals like those on display this week.
More news below.
David Meyer
Want to send thoughts or suggestions to Data Sheet? Drop a line here.
