AI failed to properly patch software flaws 74% of the time, 1Password's study warns (zdnet.com)

We might be eager to thread AI into every area of our cybersecurity defenses, but new research reveals why we should pull back.