the consensual assumption in netsec is that both red and blue teams have access to the same tools and talent pool
that's why it's a never-ending job where it's impossible to keep the edge for long
at the end of day it's a matter of means, where you can even buy zero days and use godly amounts of compute for some targets. Which is why/how state-sponsored agencies are in a different league, for instance.
the race will never end, but generally what we know is that a harder environment globally is also harder to breach meaningfully / at scale. HTTPS works, the world didn't end. you'd rather have every single house extremely hard to break than a single point of national security failure for instance. security should be in-depth vertically and decentralized horizontally (zero-trust harden every single node).
it's far from a solved topic with AI, robotics, biotech, and whatnot; but since there are way more good guys than bad guys (85-15 ratio in humanity, give or take), you'd rather ensure the 85% have all the means at their disposal, because then the means/number asymmetry is overwhelming for bad actors, who must necessarily play as underdogs. this greatly limits the amount of harm they can inflict in the long run.