AI just found the vulnerabilities, but now the people fixing them need to call in an extra thousand. Weren’t you supposed to save us some work? Bank of America asked AI to help identify software security vulnerabilities, using models like Mythos from AI company Anthropic. According to the bank’s CEO, the model uncovered a ton of issues, so the bank wanted to fix them ASAP. In Q2, they brought in 1,000 outsourced engineers to assist the existing team of about 3,000 employees. This is kind of awkward: it used to be hard to find problems, but now that AI finds them quickly, someone still has to step in to fix them. If I were sitting at a desk fixing bugs, watching AI constantly add stuff to the to-do list, I’d probably lose that brief sense of relief real quick. Sure, it’s great to catch risks early, but the people doing the work are probably going to be swamped for a while.