Two threads of the same story are pulling in opposite directions.
On one side, momentum. Brandon Tseng, cofounder of defense company Shield AI, argues the US military needs "more affordable, intelligent drones," and believes "millions" of them could be built under the Trump administration, according to a Business Insider report surfaced via Bing News. Tseng frames cheap, autonomous aircraft as something that could transform how the US fights — a shift away from small numbers of expensive platforms toward mass and machine decision-making.
On the other side, the security bill for that shift. EE Times reports that rapid military adoption of AI is bringing critical security risks with it, leaving autonomous battlefield systems vulnerable to cyberattacks. The concern isn't hypothetical hardware failure — it's that AI agents making decisions in combat become high-value targets for adversaries who want to manipulate rather than destroy them.
The tension is straightforward. The case for affordable autonomous drones rests on building them in enormous numbers. But every additional networked, software-driven system is another thing that has to be defended, patched, and trusted. Scale is the selling point and the exposure at the same time.
Neither source describes a specific breach, and the two items come from different reporting streams — one an executive's outlook on procurement, the other an assessment of security posture. Read together, they sketch a procurement push moving faster than the security thinking around it.
Why it matters: if militaries field autonomous systems by the million before the cybersecurity questions are settled, an adversary may not need to shoot anything down — they may only need to convince it to make the wrong call.