美国银行部署4000人修复AI发现的漏洞
美国银行雇佣4000人修复ai模型如anthropic的mythos和openai的daybreak发现的软件漏洞。这项工作花费数百万美元。
bank of america has dedicated 4000 people, including 1000 outside contractors and 3000 internal employees, to address software vulnerabilities identified by ai models such as anthropic's mythos and openai's daybreak. the bank has already spent a few hundred million dollars in the third quarter on this initiative, combined with anti-money-laundering requirements. ceo brian moynihan expects patching costs to decrease over time. this effort makes bank of america one of the first major us companies to disclose the scale of its response to ai-discovered security threats. the move highlights the growing importance of addressing ai-identified vulnerabilities.