This is arguably good military strategy, too - the US has ample resources to allocate to finding novel or even one-off exploits on an ongoing basis, where some poorer powers might want to rely more on keeping secret existing exploits for a longer period. Perpetually making enemies' weapons obsolete.
Also, the US has a much larger attackable surface area and far more to lose in this domain. So it makes perfect, cynical sense if the US government wants internet security in general to be better.
Just like a naval power with strong international trade interests has reasons to keep shipping lanes open to all and to deter naval piracy.
None of this requires benevolence, it's all self-interested
The naval analogy is interesting. One of the six missions of a global naval presence is to protect the sea lanes of communication. Even if that means the North Koreans occasionally manage to acquire some centrifuge tubes, the global value of trade outweighs the costs.
By very direct corollary, one would hope the NSA regards protection of the digital lanes of communication open as one of its principle missions. Trade begets economic interdependence. It's tough to go to war with major trade partners.
What's a "novel" or "one-off" exploit? I don't see the realistic dividing line between "NOBUS" exploits that NSA could realistically believe it has a proprietary grasp on and the kinds of exploits that get deployed at Pwn2Own.
Also, the US has a much larger attackable surface area and far more to lose in this domain. So it makes perfect, cynical sense if the US government wants internet security in general to be better.
Just like a naval power with strong international trade interests has reasons to keep shipping lanes open to all and to deter naval piracy.
None of this requires benevolence, it's all self-interested