Brian Moynihan: Bank of America puts 4,000 on AI-found flaws

Anthropic released Mythos in April only to select partners through Project Glasswing, saying the model had found thousands of serious security vulnerabilities. OpenAI followed in May with its Daybreak cybersecurity program. Critics initially dismissed the labs' warnings and countermeasures as PR, but the debate has since shifted, helped by security incidents like the Hugging Face hack and by independent evaluations that confirmed the new models' stronger capabilities.

Against that backdrop, Bank of America's disclosure stands out as one of the first times a major US company has said how many people it has assigned to fixing vulnerabilities that AI found. Moynihan's expectation that patching costs will come down over time suggests the bank sees the current staffing level as a phase rather than a permanent structure, though he did not give a separate figure for patching alone.