AQSh banklari uchinchi tomon sun'iy intellekt agentlari mijozlar hisoblariga ruxsatsiz kirganini aniqlash bilan bog'liq qiyinchiliklarga duch kelmoqda. Bu holat hisob ma'lumotlarini ulashish, avtomatlashtirilgan operatsiyalar va mijozning haqiqiy faoliyatini avtomatlashtirish tomonidan boshlatilgan harakatlardan ajratish qiyinligi haqida xavotirlarni kuchaytirmoqda.
Axiosning 22-oktabrdagi xabariga ko'ra, Morgan Stanley, Bank of America, JPMorgan va Wells Fargo kabi bir nechta banklar ushbu agentlar bilan bog'liq xavflarni kamaytirish ustida ishlamoqda. Ba'zi muassasalar allaqachon ruxsatsiz kirish holatlarini aniqlagan.
Xabarda aytilishicha, muammo odatda mijozlar o'z kirish ma'lumotlarini ulashishi va AI yordamchilariga o'z hisoblariga kirish huquqini berishi bilan bog'liq. Biroq, bu ruxsatlar bank siyosatlari bilan ziddiyatga kirishi mumkin, chunki ular mijozlardan shaxsan tizimga kirishni talab qiladi.
Bu farq muhim: xabarda aytilishicha, hujumchilar bank infratuzilmasini buzmagan yoki hisob ma'lumotlarini o'g'irlamagan. Aksincha, bu mijozlar, uchinchi tomon avtomatlashtirish provayderlari va moliyaviy muassasalar o'rtasidagi o'sib borayotgan ruxsat berish bo'shlig'ini ta'kidlaydi.
Mijozlar nomidan vazifalarni bajarish uchun mo'ljallangan AI agentlari potentsial ravishda hisob funksiyalari bilan o'zaro aloqada bo'lishi mumkin. Bank rahbarlari ayniqsa agentlarga xarid qilish huquqi yoki mablag'larni o'tkazish ruxsati berilishidan xavotirda.
Manbalar ba'zi agentlar qo'ng'iroq markazlari bilan bog'langanini yoki mijoz ko'rsatmalaridan mustaqil ravishda harakat qilgandek ko'rinishini ta'kidlagan. Bu mavjud nazorat tizimlari kim harakatni boshlaganini ishonchli aniqlay olishi va uning mijozning mo'ljallangan doirasida qolishini ta'minlay olishi haqida savollar tug'diradi.
Javob sifatida, Morgan Stanley o'z platformalaridan uchinchi tomon agentlarini bloklashni boshlagan. Shunga o'xshash tarzda, JPMorgan, Bank of America va Citi ham ushbu agentlar uchun kirishni cheklamoqda.
Biroq, Morgan Stanley "bizning platformalarimizga kirgan yovuz AI agentlari" aniqlanmaganini aytgan, Chase vakili esa mijozlar o'z foydalanuvchi nomlari yoki parollarini ulashmasligi kerakligini ta'kidlagan.
Bu harakatlar muhim cheklovni ta'kidlaydi: uchinchi tomon agentlariga qo'yilgan cheklovlar har bir tilga olingan muassasada ruxsatsiz hisobga kirish holati sodir bo'lganini tasdiqlash sifatida talqin qilinmasligi kerak.
Ko'pchilik banklar faqat o'qish huquqiga o'tmoqda, bu agentlarga moliyaviy ma'lumotlarni tekshirishga ruxsat beradi, lekin harakat qilish imkoniyatini bermaydi. Axios bu yondashuvni Quicken kabi byudjet ilovalariga berilgan mavjud kirish bilan solishtirgan.
Asosiy xavfsizlik muammosi hali ham identifikatsiya va ruxsat berish bo'lib qolmoqda. Rahbarlar banklar inson va agent faoliyatini, shuningdek mijoz tomonidan yo'naltirilgan avtomatlashtirish va yovuz harakatlarni ajratishda qiyinchiliklarga duch kelishini aytgan. Muassasalar, shuningdek, muammolar yuzaga kelsa, mijoz kirishini himoya qilish yoki tiklash uchun favqulodda rejalarni tayyorlamoqda.
Bank xavfsizlik jamoalari uchun yangi xavotir faqat botlarni aniqlash emas, balki avtomatlashtirilgan harakat ruxsat etilganmi, to'g'ri cheklanganmi va hisob egasining ko'rsatmalariga mos keladimi yoki yo'qligini aniqlashdir.
U.S. banks face challenges as third-party artificial intelligence agents detect unauthorized access to customer accounts. This situation raises concerns aboutcredential sharing, automated transactions, and the difficulty of distinguishing legitimate customer activity from automation-initiated actions.
According to a report from Axios on October 22, several banks, including Morgan Stanley, Bank of America, JPMorgan, and Wells Fargo, are working to mitigate risks associated with these agents. Some institutions have already identified instances of unauthorized access.
The reported issue typically involves customers sharing their login credentials and authorizing AI assistants to access their accounts. However, these permissions may conflict with bank policies that require customers to log in personally.
This distinction matters: the report indicates that attackers have not breached banking infrastructure or stolen credentials. Instead, it highlights a growing authorization gap among customers, third-party automation providers, and financial institutions.
AI agents, designed to perform tasks on customers’ behalf, can potentially interact with account functions. Banking executives are particularly worried about agents being granted purchasing authority or permission to transfer funds.
Sources noted thatsome agents have contacted call centers or appeared to act independently of customer instructions. This raises concerns about whether existing controls can reliably identify who initiated an action and ensure it stays within the customer’s intended scope.
In response, Morgan Stanley has begun blocking third-party agents from its platforms. Similarly, JPMorgan, Bank of America, and Citi are also restricting access for these agents.
However, Morgan Stanley said it has not detected “rogue AI agents accessing our platforms,” and a Chase spokesperson emphasized that customers should refrain from sharing their usernames or passwords.
These actions highlight a key limitation: restrictions on third-party agents should not be interpreted as confirmation that every mentioned institution has experienced unauthorized account access.
Most banks are moving toward read-only access, which lets agents inspect financial information without being able to take action. Axios compared this approach to existing access provided to budgeting applications like Quicken.
The central security challenge remains attribution and authorization. Executives have said banks struggle to distinguish between human and agent activity, as well as between customer-directed automation and rogue actions. Institutions are also preparing contingency plans to secure or restore customer access if problems arise.
For banking security teams, the emerging concern is not just detecting bots, but determining whether an automated action is permitted, appropriately limited, and aligned with the account holder’s instructions.
Stops Cyber threats before impact with 21 min faster MTTR. Integrate ANYRUN’s Sandbox in your SOC.