{"id":17213,"date":"2025-05-05T11:58:33","date_gmt":"2025-05-05T04:58:33","guid":{"rendered":"https:\/\/base.vn\/blog\/?p=17213"},"modified":"2025-09-05T11:19:52","modified_gmt":"2025-09-05T04:19:52","slug":"ai-trong-ngan-hang","status":"publish","type":"post","link":"https:\/\/base.vn\/blog\/ai-trong-ngan-hang\/","title":{"rendered":"AI trong ng\u00e0nh ng\u00e2n h\u00e0ng: C\u00e1ch c\u00f4ng ngh\u1ec7 thay \u0111\u1ed5i cu\u1ed9c ch\u01a1i"},"content":{"rendered":"\n

T\u1eeb vi\u1ec7c r\u00fat ng\u1eafn th\u1eddi gian x\u1eed l\u00fd giao d\u1ecbch, ng\u0103n ch\u1eb7n gian l\u1eadn t\u00e0i ch\u00ednh, \u0111\u1ebfn \u0111\u00e1nh gi\u00e1 \u0111i\u1ec3m t\u00edn d\u1ee5ng v\u00e0 c\u00e1 nh\u00e2n h\u00f3a tr\u1ea3i nghi\u1ec7m kh\u00e1ch h\u00e0ng, tr\u00ed tu\u1ec7 nh\u00e2n t\u1ea1o (AI) \u0111ang t\u00e1i \u0111\u1ecbnh h\u00ecnh ng\u00e0nh ng\u00e2n h\u00e0ng tr\u00ean quy m\u00f4 to\u00e0n c\u1ea7u. Vi\u1ec7t Nam c\u0169ng kh\u00f4ng n\u1eb1m ngo\u00e0i xu th\u1ebf \u0111\u00f3, khi ng\u00e0y c\u00e0ng nhi\u1ec1u t\u1ed5 ch\u1ee9c t\u00e0i ch\u00ednh t\u00edch c\u1ef1c \u1ee9ng d\u1ee5ng AI \u0111\u1ec3 t\u1ed1i \u01b0u hi\u1ec7u qu\u1ea3 v\u1eadn h\u00e0nh v\u00e0 n\u00e2ng cao n\u0103ng l\u1ef1c c\u1ea1nh tranh.<\/p>\n\n\n\n

V\u1eady c\u1ee5 th\u1ec3, vi\u1ec7c \u1ee9ng d\u1ee5ng AI trong ng\u00e2n h\u00e0ng l\u00e0 nh\u01b0 th\u1ebf n\u00e0o? Nh\u1eefng t\u1ed5 ch\u1ee9c n\u00e0o \u0111ang d\u1eabn \u0111\u1ea7u v\u00e0 t\u1ea1o ra nh\u1eefng \u0111\u1ed9t ph\u00e1 \u0111\u00e1ng k\u1ec3 nh\u1edd c\u00f4ng ngh\u1ec7 n\u00e0y? H\u00e3y c\u00f9ng Base.vn<\/a> kh\u00e1m ph\u00e1 to\u00e0n c\u1ea3nh b\u1ee9c tranh \u1ee9ng d\u1ee5ng AI trong ng\u00e0nh ng\u00e2n h\u00e0ng, t\u1eeb c\u00e1c xu h\u01b0\u1edbng n\u1ed5i b\u1eadt tr\u00ean th\u1ebf gi\u1edbi \u0111\u1ebfn hi\u1ec7n tr\u1ea1ng v\u00e0 ti\u1ec1m n\u0103ng t\u1ea1i Vi\u1ec7t Nam.<\/p>\n\n\n\n

1. \u1ee8ng d\u1ee5ng AI trong ng\u00e2n h\u00e0ng l\u00e0 nh\u01b0 th\u1ebf n\u00e0o?<\/strong><\/h2>\n\n\n\n

AI trong ng\u00e2n h\u00e0ng<\/strong> \u0111\u1ec1 c\u1eadp \u0111\u1ebfn vi\u1ec7c \u1ee9ng d\u1ee5ng c\u00e1c c\u00f4ng ngh\u1ec7 tr\u00ed tu\u1ec7 nh\u00e2n t\u1ea1o, \u0111\u1eb7c bi\u1ec7t l\u00e0 h\u1ecdc m\u00e1y (Machine Learning) v\u00e0 x\u1eed l\u00fd ng\u00f4n ng\u1eef t\u1ef1 nhi\u00ean (Natural Processing Language), \u0111\u1ec3 ph\u00e2n t\u00edch d\u1eef li\u1ec7u s\u00e2u h\u01a1n, t\u1ef1 \u0111\u1ed9ng h\u00f3a t\u00e1c v\u1ee5 v\u00e0 h\u1ed7 tr\u1ee3 ra quy\u1ebft \u0111\u1ecbnh m\u1ed9t c\u00e1ch nhanh ch\u00f3ng, ch\u00ednh x\u00e1c. Qua \u0111\u00f3, c\u00e1c ng\u00e2n h\u00e0ng v\u00e0 t\u1ed5 ch\u1ee9c t\u00e0i ch\u00ednh c\u00f3 th\u1ec3 t\u0103ng t\u1ed1c \u0111\u1ed9 x\u1eed l\u00fd giao d\u1ecbch, n\u00e2ng cao hi\u1ec7u qu\u1ea3 qu\u1ea3n l\u00fd r\u1ee7i ro, ph\u00e1t hi\u1ec7n v\u00e0 ng\u0103n ch\u1eb7n gian l\u1eadn k\u1ecbp th\u1eddi, c\u1ea3i thi\u1ec7n t\u01b0\u01a1ng t\u00e1c v\u1edbi kh\u00e1ch h\u00e0ng, v\u00e0 t\u1ed1i \u01b0u h\u00f3a v\u1eadn h\u00e0nh \u1edf nhi\u1ec1u c\u1ea5p \u0111\u1ed9.<\/p>\n\n\n\n

Nh\u00ecn chung, c\u00f4ng ngh\u1ec7 AI<\/a> \u0111ang m\u1edf ra m\u1ed9t k\u1ef7 nguy\u00ean m\u1edbi cho ng\u00e0nh t\u00e0i ch\u00ednh \u2013 ng\u00e2n h\u00e0ng, n\u01a1i m\u1ecdi quy\u1ebft \u0111\u1ecbnh \u0111\u1ec1u d\u1ef1a tr\u00ean d\u1eef li\u1ec7u theo th\u1eddi gian th\u1ef1c, \u0111em \u0111\u1ebfn cho kh\u00e1ch h\u00e0ng tr\u1ea3i nghi\u1ec7m an to\u00e0n v\u00e0 thu\u1eadn ti\u1ec7n h\u01a1n bao gi\u1edd h\u1ebft.<\/p>\n\n\n

\n
\"AI<\/figure><\/div>\n\n\n

2. L\u1ee3i \u00edch, \u1ee9ng d\u1ee5ng c\u1ee7a AI trong ng\u00e0nh ng\u00e2n h\u00e0ng v\u00e0 v\u00ed d\u1ee5 th\u1ef1c t\u1ebf<\/strong><\/h2>\n\n\n\n

Hi\u1ec7n nay, tr\u00ed tu\u1ec7 nh\u00e2n t\u1ea1o \u0111\u00e3 v\u00e0 \u0111ang \u0111\u01b0\u1ee3c \u1ee9ng d\u1ee5ng r\u1ed9ng r\u00e3i trong l\u0129nh v\u1ef1c t\u00e0i ch\u00ednh \u2013 ng\u00e2n h\u00e0ng v\u00e0 m\u1edf ra ti\u1ec1m n\u0103ng ch\u01b0a c\u00f3 gi\u1edbi h\u1ea1n. D\u01b0\u1edbi \u0111\u00e2y l\u00e0 m\u1ed9t s\u1ed1 \u1ee9ng d\u1ee5ng ti\u00eau bi\u1ec3u c\u00f9ng v\u00ed d\u1ee5 c\u1ee5 th\u1ec3 gi\u00fap minh h\u1ecda r\u00f5 n\u00e9t vai tr\u00f2 c\u1ee7a AI trong ng\u00e0nh:<\/p>\n\n\n\n

2.1 C\u1ea3i thi\u1ec7n d\u1ecbch v\u1ee5 kh\u00e1ch h\u00e0ng v\u1edbi tr\u1ee3 l\u00fd \u1ea3o AI<\/strong><\/h3>\n\n\n\n

AI \u0111\u00e3 g\u00f3p ph\u1ea7n thay \u0111\u1ed5i c\u00e1ch ng\u00e2n h\u00e0ng ph\u1ee5c v\u1ee5 kh\u00e1ch h\u00e0ng, \u0111\u1eb7c bi\u1ec7t th\u00f4ng qua c\u00e1c tr\u1ee3 l\u00fd \u1ea3o v\u00e0 chatbot. Nh\u1eefng c\u00f4ng c\u1ee5 n\u00e0y c\u00f3 th\u1ec3 h\u1ed7 tr\u1ee3 kh\u00e1ch h\u00e0ng li\u00ean t\u1ee5c 24\/7, k\u1ec3 c\u1ea3 ngo\u00e0i gi\u1edd h\u00e0nh ch\u00ednh, cu\u1ed1i tu\u1ea7n hay ng\u00e0y l\u1ec5, \u0111i\u1ec1u m\u00e0 c\u00e1c m\u00f4 h\u00ecnh d\u1ecbch v\u1ee5 truy\u1ec1n th\u1ed1ng kh\u00f4ng th\u1ec3 \u0111\u00e1p \u1ee9ng \u0111\u01b0\u1ee3c.<\/p>\n\n\n\n

M\u1ed9t v\u00ed d\u1ee5 n\u1ed5i b\u1eadt l\u00e0 \u201cErica\u201d c\u1ee7a Bank of America \u2013 m\u1ed9t tr\u1ee3 l\u00fd \u1ea3o h\u1ed7 tr\u1ee3 kh\u00e1ch h\u00e0ng th\u1ef1c hi\u1ec7n c\u00e1c t\u00e1c v\u1ee5 nh\u01b0 ki\u1ec3m tra s\u1ed1 d\u01b0, thanh to\u00e1n h\u00f3a \u0111\u01a1n, hay nh\u1eadn t\u01b0 v\u1ea5n t\u00e0i ch\u00ednh c\u00e1 nh\u00e2n. Nh\u1edd c\u00f4ng ngh\u1ec7 x\u1eed l\u00fd ng\u00f4n ng\u1eef t\u1ef1 nhi\u00ean (NLP) v\u00e0 h\u1ecdc m\u00e1y (ML), Erica c\u00f3 th\u1ec3 hi\u1ec3u v\u00e0 ph\u1ea3n h\u1ed3i c\u00e2u h\u1ecfi gi\u1ed1ng nh\u01b0 con ng\u01b0\u1eddi. T\u00ednh \u0111\u1ebfn nay, Erica \u0111\u00e3 ghi nh\u1eadn h\u01a1n 1,5 t\u1ef7 l\u01b0\u1ee3t t\u01b0\u01a1ng t\u00e1c<\/a> k\u1ec3 t\u1eeb khi ra m\u1eaft v\u00e0o n\u0103m 2018.<\/p>\n\n\n\n

\u201cFargo\u201d c\u1ee7a Wells Fargo, c\u0169ng l\u00e0 m\u1ed9t tr\u1ee3 l\u00fd \u1ea3o \u1ee9ng d\u1ee5ng AI th\u00f4ng minh, \u0111\u01b0\u1ee3c x\u00e2y d\u1ef1ng d\u1ef1a tr\u00ean Google Dialogflow v\u00e0 m\u00f4 h\u00ecnh ng\u00f4n ng\u1eef l\u1edbn PaLM 2. Fargo c\u00f3 th\u1ec3 x\u1eed l\u00fd m\u01b0\u1ee3t m\u00e0 c\u00e1c y\u00eau c\u1ea7u nh\u01b0 chuy\u1ec3n ti\u1ec1n, thanh to\u00e1n h\u00f3a \u0111\u01a1n b\u1eb1ng gi\u1ecdng n\u00f3i ho\u1eb7c v\u0103n b\u1ea3n, v\u1edbi trung b\u00ecnh 2,7 l\u01b0\u1ee3t t\u01b0\u01a1ng t\u00e1c m\u1ed7i phi\u00ean<\/a> ho\u1ea1t \u0111\u1ed9ng.<\/p>\n\n\n\n

V\u1ec1 m\u1eb7t k\u1ef9 thu\u1eadt, \u0111\u1ec3 ph\u00e1t tri\u1ec3n th\u00e0nh c\u00f4ng nh\u1eefng c\u00f4ng c\u1ee5 n\u00e0y, c\u00e1c ng\u00e2n h\u00e0ng ph\u1ea3i hu\u1ea5n luy\u1ec7n h\u1ec7 th\u1ed1ng tr\u00ean l\u01b0\u1ee3ng d\u1eef li\u1ec7u l\u1edbn, \u0111a d\u1ea1ng v\u1ec1 gi\u1ecdng n\u00f3i v\u00e0 ph\u01b0\u01a1ng ng\u1eef, \u0111\u1ed3ng th\u1eddi t\u1ed1i \u01b0u h\u00f3a kh\u1ea3 n\u0103ng k\u1ebft n\u1ed1i v\u1edbi c\u00e1c h\u1ec7 th\u1ed1ng n\u1ed9i b\u1ed9 qua API. Ngo\u00e0i ra, y\u1ebfu t\u1ed1 b\u1ea3o m\u1eadt c\u0169ng c\u1ea7n \u0111\u01b0\u1ee3c ch\u00fa tr\u1ecdng v\u1edbi c\u00e1c giao th\u1ee9c truy\u1ec1n th\u00f4ng an to\u00e0n v\u00e0 c\u01a1 ch\u1ebf m\u00e3 h\u00f3a v\u1eefng ch\u1eafc nh\u1eb1m b\u1ea3o v\u1ec7 th\u00f4ng tin nh\u1ea1y c\u1ea3m c\u1ee7a kh\u00e1ch h\u00e0ng.<\/p>\n\n\n\n

2.2 Ph\u00e1t hi\u1ec7n v\u00e0 ph\u00f2ng ch\u1ed1ng h\u00e0nh vi gian l\u1eadn<\/strong><\/h3>\n\n\n\n

M\u1ed9t trong nh\u1eefng \u1ee9ng d\u1ee5ng n\u1ed5i b\u1eadt kh\u00e1c c\u1ee7a AI trong ng\u00e2n h\u00e0ng l\u00e0 kh\u1ea3 n\u0103ng ph\u00e1t hi\u1ec7n v\u00e0 ng\u0103n ch\u1eb7n gian l\u1eadn giao d\u1ecbch m\u1ed9t c\u00e1ch chu\u1ea9n x\u00e1c. Th\u00f4ng qua vi\u1ec7c gi\u00e1m s\u00e1t c\u00e1c giao d\u1ecbch theo th\u1eddi gian th\u1ef1c, AI c\u00f3 th\u1ec3 ph\u00e2n t\u00edch h\u00e0nh vi ng\u01b0\u1eddi d\u00f9ng, nh\u1eadn di\u1ec7n c\u00e1c d\u1ea5u hi\u1ec7u b\u1ea5t th\u01b0\u1eddng v\u00e0 \u0111\u01b0a ra c\u1ea3nh b\u00e1o s\u1edbm \u2013 t\u1eeb \u0111\u00f3 gi\u00fap ng\u00e2n h\u00e0ng ch\u1ee7 \u0111\u1ed9ng ng\u0103n ch\u1eb7n c\u00e1c h\u00e0nh vi gian l\u1eadn tr\u01b0\u1edbc khi ch\u00fang g\u00e2y ra thi\u1ec7t h\u1ea1i. Kh\u00f4ng ch\u1ec9 gi\u00fap b\u1ea3o v\u1ec7 t\u00e0i kho\u1ea3n kh\u00e1ch h\u00e0ng, c\u00e1ch ti\u1ebfp c\u1eadn n\u00e0y c\u00f2n g\u00f3p ph\u1ea7n t\u0103ng c\u01b0\u1eddng ni\u1ec1m tin c\u1ee7a kh\u00e1ch h\u00e0ng v\u00e0o h\u1ec7 th\u1ed1ng b\u1ea3o m\u1eadt c\u1ee7a ng\u00e2n h\u00e0ng.<\/p>\n\n\n\n

Mastercard l\u00e0 m\u1ed9t trong nh\u1eefng \u0111\u01a1n v\u1ecb ti\u00ean phong trong vi\u1ec7c \u1ee9ng d\u1ee5ng AI v\u00e0o ph\u00f2ng ch\u1ed1ng gian l\u1eadn. V\u1edbi kh\u1ea3 n\u0103ng x\u1eed l\u00fd v\u00e0 ph\u00e2n t\u00edch kh\u1ed1i l\u01b0\u1ee3ng l\u1edbn d\u1eef li\u1ec7u giao d\u1ecbch theo th\u1eddi gian th\u1ef1c, h\u1ec7 th\u1ed1ng AI c\u1ee7a Mastercard \u0111\u00e3 gi\u00fap n\u00e2ng cao t\u1ed1c \u0111\u1ed9 ph\u00e1t hi\u1ec7n gian l\u1eadn l\u00ean \u0111\u1ebfn 300%<\/a>. C\u00e1c h\u1ec7 th\u1ed1ng n\u00e0y d\u1ef1a tr\u00ean m\u00f4 h\u00ecnh h\u1ecdc c\u00f3 gi\u00e1m s\u00e1t (supervised learning), \u0111\u01b0\u1ee3c hu\u1ea5n luy\u1ec7n b\u1eb1ng d\u1eef li\u1ec7u l\u1ecbch s\u1eed, trong \u0111\u00f3 c\u00e1c giao d\u1ecbch \u0111\u00e3 \u0111\u01b0\u1ee3c g\u1eafn nh\u00e3n r\u00f5 r\u00e0ng l\u00e0 \u201ch\u1ee3p l\u1ec7\u201d ho\u1eb7c \u201cgian l\u1eadn\u201d. H\u01a1n n\u1eefa, theo th\u1eddi gian, h\u1ec7 th\u1ed1ng s\u1ebd ng\u00e0y c\u00e0ng tinh vi h\u01a1n, nh\u1edd v\u00e0o kh\u1ea3 n\u0103ng h\u1ecdc h\u1ecfi v\u00e0 c\u1eadp nh\u1eadt li\u00ean t\u1ee5c t\u1eeb c\u00e1c m\u1eabu d\u1eef li\u1ec7u th\u1ef1c t\u1ebf.<\/p>\n\n\n

\n
\"L\u1ee3i<\/figure><\/div>\n\n\n

2.3 T\u1ed1i \u01b0u h\u00f3a c\u00e1c quy tr\u00ecnh n\u1ed9i b\u1ed9<\/strong><\/h3>\n\n\n\n

M\u1ed9t trong nh\u1eefng \u1ee9ng d\u1ee5ng ph\u1ed5 bi\u1ebfn nh\u1ea5t c\u1ee7a AI trong ng\u00e2n h\u00e0ng l\u00e0 t\u1ef1 \u0111\u1ed9ng h\u00f3a c\u00e1c t\u00e1c v\u1ee5 l\u1eb7p \u0111i l\u1eb7p l\u1ea1i \u0111\u1ec3 n\u00e2ng cao hi\u1ec7u qu\u1ea3 x\u1eed l\u00fd c\u00f4ng vi\u1ec7c \u1edf m\u1ecdi c\u1ea5p \u0111\u1ed9 ph\u00f2ng ban.<\/p>\n\n\n\n

L\u1ea5y ng\u00e2n h\u00e0ng OCBC l\u00e0m h\u00ecnh m\u1eabu. T\u1ed5 ch\u1ee9c n\u00e0y \u0111\u00e3 tri\u1ec3n khai \u201cOCBC GPT\u201d \u2013 m\u1ed9t chatbot n\u1ed9i b\u1ed9 t\u00edch h\u1ee3p AI, h\u1ed7 tr\u1ee3 nh\u00e2n vi\u00ean trong nhi\u1ec1u c\u00f4ng vi\u1ec7c th\u01b0\u1eddng nh\u1eadt nh\u01b0 so\u1ea1n th\u1ea3o t\u00e0i li\u1ec7u, t\u00ecm ki\u1ebfm th\u00f4ng tin v\u00e0 nghi\u00ean c\u1ee9u ki\u1ebfn th\u1ee9c chuy\u00ean ng\u00e0nh. Nh\u1edd c\u00f4ng c\u1ee5 n\u00e0y, nh\u00e2n vi\u00ean c\u00f3 th\u1ec3 ti\u1ebft ki\u1ec7m th\u1eddi gian cho c\u00e1c t\u00e1c v\u1ee5 \u0111\u01a1n gi\u1ea3n \u0111\u1ec3 t\u1eadp trung x\u1eed l\u00fd c\u00e1c v\u1ea5n \u0111\u1ec1 chuy\u00ean s\u00e2u v\u00e0 ph\u1ee9c t\u1ea1p h\u01a1n.<\/p>\n\n\n\n

Tuy nhi\u00ean, \u0111\u1ec3 tri\u1ec3n khai th\u00e0nh c\u00f4ng c\u00e1c gi\u1ea3i ph\u00e1p AI nh\u01b0 OCBC GPT, ng\u00e2n h\u00e0ng c\u1ea7n v\u01b0\u1ee3t qua m\u1ed9t s\u1ed1 th\u00e1ch th\u1ee9c k\u1ef9 thu\u1eadt, trong \u0111\u00f3 quan tr\u1ecdng nh\u1ea5t l\u00e0 vi\u1ec7c t\u00edch h\u1ee3p li\u1ec1n m\u1ea1ch v\u1edbi h\u1ea1 t\u1ea7ng CNTT hi\u1ec7n c\u00f3, cho ph\u00e9p nh\u00e2n vi\u00ean truy c\u1eadp d\u1ec5 d\u00e0ng v\u00e0o d\u1eef li\u1ec7u c\u1ea7n thi\u1ebft trong khi v\u1eabn \u0111\u1ea3m b\u1ea3o t\u00ednh b\u1ea3o m\u1eadt v\u00e0 quy\u1ec1n ri\u00eang t\u01b0 cao.<\/p>\n\n\n\n

Ngo\u00e0i ra, c\u00e1c chatbot AI th\u01b0\u1eddng ph\u1ea3i \u0111\u01b0\u1ee3c t\u00f9y ch\u1ec9nh theo d\u1eef li\u1ec7u n\u1ed9i b\u1ed9, nh\u1eb1m x\u1eed l\u00fd linh ho\u1ea1t c\u00e1c quy tr\u00ecnh \u0111\u1eb7c th\u00f9 v\u00e0 thu\u1eadt ng\u1eef chuy\u00ean ng\u00e0nh. \u0110i\u1ec1u n\u00e0y \u0111\u00f2i h\u1ecfi s\u1ef1 gi\u00e1m s\u00e1t, tinh ch\u1ec9nh v\u00e0 hu\u1ea5n luy\u1ec7n li\u00ean t\u1ee5c. Cu\u1ed1i c\u00f9ng, giao di\u1ec7n c\u1ee7a c\u00f4ng c\u1ee5 c\u1ea7n \u0111\u01b0\u1ee3c thi\u1ebft k\u1ebf tr\u1ef1c quan, d\u1ec5 s\u1eed d\u1ee5ng v\u00e0 \u0111\u00f4i khi c\u1ea7n \u0111\u1ebfn c\u00e1c ti\u1ec7n \u00edch m\u1edf r\u1ed9ng ho\u1eb7c plugin t\u00f9y ch\u1ec9nh \u0111\u1ec3 \u0111\u00e1p \u1ee9ng nhu c\u1ea7u th\u1ef1c t\u1ebf.<\/p>\n\n\n\n

2.4 \u0110\u00e1nh gi\u00e1 \u0111i\u1ec3m t\u00edn d\u1ee5ng c\u00f4ng b\u1eb1ng v\u00e0 to\u00e0n di\u1ec7n h\u01a1n<\/strong><\/h3>\n\n\n\n

Theo ph\u01b0\u01a1ng ph\u00e1p truy\u1ec1n th\u1ed1ng, \u0111i\u1ec3m t\u00edn d\u1ee5ng ch\u1ee7 y\u1ebfu \u0111\u01b0\u1ee3c \u0111\u00e1nh gi\u00e1 d\u1ef1a tr\u00ean c\u00e1c y\u1ebfu t\u1ed1 nh\u01b0 l\u1ecbch s\u1eed thanh to\u00e1n, h\u1ea1n m\u1ee9c t\u00edn d\u1ee5ng v\u00e0 c\u00e1c l\u1ea7n ki\u1ec3m tra t\u00edn d\u1ee5ng g\u1ea7n nh\u1ea5t. Tuy nhi\u00ean, c\u00e1c h\u1ec7 th\u1ed1ng t\u00edch h\u1ee3p AI \u0111\u00e3 mang \u0111\u1ebfn m\u1ed9t c\u00e1ch ti\u1ebfp c\u1eadn ho\u00e0n to\u00e0n m\u1edbi \u2013 ph\u00e2n t\u00edch d\u1ef1a tr\u00ean thu\u1eadt to\u00e1n h\u1ecdc m\u00e1y (ML)<\/a>, cho ph\u00e9p ng\u00e2n h\u00e0ng x\u1eed l\u00fd kh\u1ed1i l\u01b0\u1ee3ng l\u1edbn d\u1eef li\u1ec7u th\u1eddi gian th\u1ef1c, t\u1eeb \u0111\u00f3 x\u00e1c \u0111\u1ecbnh c\u00e1c m\u00f4 h\u00ecnh h\u00e0nh vi, \u0111\u01b0a ra d\u1ef1 \u0111o\u00e1n ch\u00ednh x\u00e1c, c\u00f4ng b\u1eb1ng v\u00e0 c\u00e1 nh\u00e2n h\u00f3a h\u01a1n.<\/p>\n\n\n\n

AI c\u00f2n c\u00f3 kh\u1ea3 n\u0103ng khai th\u00e1c c\u00e1c \u0111i\u1ec3m d\u1eef li\u1ec7u phi truy\u1ec1n th\u1ed1ng \u2013 ch\u1eb3ng h\u1ea1n nh\u01b0 ho\u1ea1t \u0111\u1ed9ng m\u1ea1ng x\u00e3 h\u1ed9i ho\u1eb7c h\u00e0nh vi mua s\u1eafm tr\u1ef1c tuy\u1ebfn \u2013 v\u1ed1n th\u01b0\u1eddng b\u1ecb b\u1ecf qua trong c\u00e1c m\u00f4 h\u00ecnh \u0111\u00e1nh gi\u00e1 t\u00edn d\u1ee5ng c\u0169. Nh\u1edd \u0111\u00f3, ng\u00e2n h\u00e0ng c\u00f3 th\u1ec3 hi\u1ec3u r\u00f5 h\u01a1n v\u1ec1 th\u00f3i quen t\u00e0i ch\u00ednh v\u00e0 m\u1ee9c \u0111\u1ed9 uy t\u00edn c\u1ee7a ng\u01b0\u1eddi vay, \u0111\u1eb7c bi\u1ec7t l\u00e0 nh\u1eefng ng\u01b0\u1eddi ch\u01b0a c\u00f3 l\u1ecbch s\u1eed t\u00edn d\u1ee5ng ho\u1eb7c c\u00f3 h\u1ed3 s\u01a1 t\u00edn d\u1ee5ng c\u00f2n h\u1ea1n ch\u1ebf.<\/p>\n\n\n\n

M\u1ed9t v\u00ed d\u1ee5 n\u1ed5i b\u1eadt l\u00e0 Upstart \u2013 n\u1ec1n t\u1ea3ng cho vay \u1ee9ng d\u1ee5ng AI \u0111\u1ec3 \u0111\u00e1nh gi\u00e1 r\u1ee7i ro v\u00e0 x\u00e1c \u0111\u1ecbnh \u0111i\u1ec1u ki\u1ec7n vay. Thay v\u00ec ch\u1ec9 d\u1ef1a v\u00e0o \u0111i\u1ec3m t\u00edn d\u1ee5ng truy\u1ec1n th\u1ed1ng, Upstart ph\u00e2n t\u00edch th\u00eam nhi\u1ec1u y\u1ebfu t\u1ed1 kh\u00e1c nh\u01b0 tr\u00ecnh \u0111\u1ed9 h\u1ecdc v\u1ea5n, l\u1ecbch s\u1eed v\u00e0 kinh nghi\u1ec7m l\u00e0m vi\u1ec7c, mang \u0111\u1ebfn m\u1ed9t c\u00e1i nh\u00ecn to\u00e0n di\u1ec7n h\u01a1n v\u1ec1 kh\u00e1ch h\u00e0ng. Nh\u1edd ph\u01b0\u01a1ng ph\u00e1p n\u00e0y, Upstart kh\u00f4ng ch\u1ec9 m\u1edf r\u1ed9ng kh\u1ea3 n\u0103ng ti\u1ebfp c\u1eadn t\u00edn d\u1ee5ng cho nh\u1eefng ng\u01b0\u1eddi m\u1edbi t\u1ed1t nghi\u1ec7p ho\u1eb7c ch\u01b0a c\u00f3 l\u1ecbch s\u1eed vay, m\u00e0 c\u00f2n gi\u1ea3m t\u1ef7 l\u1ec7 v\u1ee1 n\u1ee3 \u0111\u1ebfn 75%<\/a>.<\/p>\n\n\n\n

2.5 C\u00e1 nh\u00e2n h\u00f3a khuy\u1ebfn ngh\u1ecb s\u1ea3n ph\u1ea9m ho\u1eb7c d\u1ecbch v\u1ee5<\/strong><\/h3>\n\n\n\n

Th\u00f4ng qua vi\u1ec7c ph\u00e2n t\u00edch d\u1eef li\u1ec7u h\u00e0nh vi v\u00e0 l\u1ecbch s\u1eed giao d\u1ecbch, c\u00e1c thu\u1eadt to\u00e1n AI c\u00f3 th\u1ec3 x\u00e1c \u0111\u1ecbnh th\u00f3i quen chi ti\u00eau, nhu c\u1ea7u t\u00e0i ch\u00ednh v\u00e0 xu h\u01b0\u1edbng \u0111\u1ea7u t\u01b0 c\u1ee7a t\u1eebng c\u00e1 nh\u00e2n. T\u1eeb \u0111\u00f3, ng\u00e2n h\u00e0ng c\u00f3 th\u1ec3 \u0111\u01b0a ra c\u00e1c g\u1ee3i \u00fd v\u00e0 l\u1eddi khuy\u00ean ph\u00f9 h\u1ee3p v\u1edbi t\u1eebng kh\u00e1ch h\u00e0ng, gi\u00fap h\u1ecd \u0111\u01b0a ra quy\u1ebft \u0111\u1ecbnh t\u00e0i ch\u00ednh s\u00e1ng su\u1ed1t h\u01a1n.<\/p>\n\n\n\n

S\u1ef1 c\u00e1 nh\u00e2n h\u00f3a n\u00e0y kh\u00f4ng ch\u1ec9 n\u00e2ng cao tr\u1ea3i nghi\u1ec7m ng\u01b0\u1eddi d\u00f9ng m\u00e0 c\u00f2n t\u0103ng c\u01b0\u1eddng m\u1ee9c \u0111\u1ed9 h\u00e0i l\u00f2ng v\u00e0 l\u00f2ng trung th\u00e0nh v\u1edbi th\u01b0\u01a1ng hi\u1ec7u. Khi kh\u00e1ch h\u00e0ng c\u1ea3m nh\u1eadn \u0111\u01b0\u1ee3c s\u1ef1 th\u1ea5u hi\u1ec3u v\u00e0 \u0111\u1ed3ng h\u00e0nh t\u1eeb ng\u00e2n h\u00e0ng, h\u1ecd c\u00f3 xu h\u01b0\u1edbng g\u1eafn b\u00f3 l\u00e2u d\u00e0i h\u01a1n.<\/p>\n\n\n\n

M\u1ed9t v\u00ed d\u1ee5 \u0111i\u1ec3n h\u00ecnh l\u00e0 Glass \u2013 n\u1ec1n t\u1ea3ng ph\u00e2n t\u00edch h\u1ed7 tr\u1ee3 AI c\u1ee7a Bank of America. Glass k\u1ebft h\u1ee3p gi\u1eefa d\u1eef li\u1ec7u th\u1ecb tr\u01b0\u1eddng v\u00e0 c\u00e1c m\u00f4 h\u00ecnh t\u00e0i ch\u00ednh n\u1ed9i b\u1ed9, s\u1eed d\u1ee5ng k\u1ef9 thu\u1eadt h\u1ecdc m\u00e1y \u0111\u1ec3 d\u1ef1 \u0111o\u00e1n xu h\u01b0\u1edbng ng\u00e0nh v\u00e0 nhu c\u1ea7u kh\u00e1ch h\u00e0ng. Nh\u1edd \u0111\u00f3, ng\u00e2n h\u00e0ng c\u00f3 th\u1ec3 \u0111\u01b0a ra c\u00e1c khuy\u1ebfn ngh\u1ecb \u0111\u1ea7u t\u01b0 mang t\u00ednh c\u00e1 nh\u00e2n h\u00f3a cao, \u0111\u1ed3ng th\u1eddi th\u1ec3 hi\u1ec7n vai tr\u00f2 ti\u00ean phong trong vi\u1ec7c \u1ee9ng d\u1ee5ng AI v\u00e0o l\u0129nh v\u1ef1c t\u01b0 v\u1ea5n t\u00e0i ch\u00ednh..<\/p>\n\n\n

\n
\"L\u1ee3i<\/figure><\/div>\n\n\n

2.6 Th\u00fac \u0111\u1ea9y ti\u1ec1m n\u0103ng s\u00e1ng t\u1ea1o c\u1ee7a Neobank v\u00e0 App Bank<\/strong><\/h3>\n\n\n\n

V\u1edbi nh\u1eefng m\u00f4 h\u00ecnh ng\u00e2n h\u00e0ng s\u1ed1 nh\u01b0 Neobank v\u00e0 App Bank (v\u1ed1n kh\u00f4ng c\u00f3 chi nh\u00e1nh hay ph\u00f2ng giao d\u1ecbch v\u1eadt l\u00fd m\u00e0 ho\u1ea1t \u0111\u1ed9ng ho\u00e0n to\u00e0n tr\u00ean m\u00f4i tr\u01b0\u1eddng s\u1ed1), AI \u0111ang m\u1edf ra nh\u1eefng h\u01b0\u1edbng \u0111i m\u1edbi m\u1ebb v\u00e0 s\u00e1ng t\u1ea1o h\u01a1n trong vi\u1ec7c c\u00e1 nh\u00e2n h\u00f3a tr\u1ea3i nghi\u1ec7m ng\u01b0\u1eddi d\u00f9ng.<\/p>\n\n\n\n

Nh\u1edd kh\u1ea3 n\u0103ng h\u1ecdc t\u1eadp t\u1eeb h\u00e0nh vi v\u00e0 t\u01b0\u01a1ng t\u00e1c c\u1ee7a ng\u01b0\u1eddi d\u00f9ng, AI c\u00f3 th\u1ec3 t\u1ef1 \u0111\u1ed9ng tinh ch\u1ec9nh giao di\u1ec7n v\u00e0 t\u00ednh n\u0103ng c\u1ee7a app ng\u00e2n h\u00e0ng sao cho ph\u00f9 h\u1ee3p v\u1edbi s\u1edf th\u00edch v\u00e0 nhu c\u1ea7u ri\u00eang c\u1ee7a t\u1eebng c\u00e1 nh\u00e2n. V\u00ed d\u1ee5, n\u1ebfu m\u1ed9t ng\u01b0\u1eddi th\u01b0\u1eddng xuy\u00ean ki\u1ec3m tra danh m\u1ee5c \u0111\u1ea7u t\u01b0, h\u1ec7 th\u1ed1ng c\u00f3 th\u1ec3 t\u1ef1 \u0111\u1ed9ng \u01b0u ti\u00ean hi\u1ec3n th\u1ecb c\u00e1c t\u00ednh n\u0103ng li\u00ean quan ngay tr\u00ean m\u00e0n h\u00ecnh ch\u00ednh c\u1ee7a app \u2013 gi\u00fap ng\u01b0\u1eddi d\u00f9ng ti\u1ebfp c\u1eadn nhanh ch\u00f3ng m\u00e0 kh\u00f4ng c\u1ea7n thao t\u00e1c th\u00eam.<\/p>\n\n\n\n

AI c\u0169ng c\u00f3 th\u1ec3 ph\u00e2n t\u00edch th\u00f3i quen s\u1eed d\u1ee5ng h\u00e0ng ng\u00e0y \u0111\u1ec3 t\u1ed1i \u01b0u h\u00f3a tr\u1ea3i nghi\u1ec7m. N\u1ebfu ng\u01b0\u1eddi d\u00f9ng th\u01b0\u1eddng th\u1ef1c hi\u1ec7n chuy\u1ec3n kho\u1ea3n v\u00e0o l\u00fac 2 gi\u1edd chi\u1ec1u m\u1ed7i ng\u00e0y, app c\u00f3 th\u1ec3 ch\u1ee7 \u0111\u1ed9ng \u0111i\u1ec1u ch\u1ec9nh giao di\u1ec7n t\u1ea1i th\u1eddi \u0111i\u1ec3m \u0111\u00f3 \u0111\u1ec3 r\u00fat ng\u1eafn quy tr\u00ecnh giao d\u1ecbch.<\/p>\n\n\n\n

B\u1eb1ng c\u00e1ch t\u1eadn d\u1ee5ng nh\u1eefng kh\u1ea3 n\u0103ng n\u00e0y c\u1ee7a AI, ng\u00e2n h\u00e0ng s\u1ed1 c\u00f3 th\u1ec3 t\u1ea1o ra m\u1ed9t tr\u1ea3i nghi\u1ec7m giao d\u1ecbch mang t\u00ednh c\u00e1 nh\u00e2n h\u00f3a v\u01b0\u1ee3t tr\u1ed9i, gi\u00fap ng\u01b0\u1eddi d\u00f9ng c\u1ea3m th\u1ea5y app \u0111\u01b0\u1ee3c thi\u1ebft k\u1ebf d\u00e0nh ri\u00eang cho m\u00ecnh \u2013 t\u1eeb \u0111\u00f3 t\u0103ng s\u1ef1 th\u00e2n thu\u1ed9c v\u00e0 n\u00e2ng cao tr\u1ea3i nghi\u1ec7m t\u1ed5ng th\u1ec3.<\/p>\n\n\n\n

\u0110\u1ecdc th\u00eam:<\/strong> Chuy\u1ec3n \u0111\u1ed5i s\u1ed1 ng\u00e0nh ng\u00e2n h\u00e0ng: Xu h\u01b0\u1edbng trong th\u1eddi \u0111\u1ea1i s\u1ed1 ho\u00e1<\/a><\/p>\n\n\n\n

3. Th\u1ef1c tr\u1ea1ng \u1ee9ng d\u1ee5ng tr\u00ed tu\u1ec7 nh\u00e2n t\u1ea1o trong ng\u00e2n h\u00e0ng t\u1ea1i Vi\u1ec7t Nam v\u00e0 tr\u00ean th\u1ebf gi\u1edbi<\/strong><\/h2>\n\n\n\n

Trong b\u1ed1i c\u1ea3nh ng\u00e0nh t\u00e0i ch\u00ednh \u2013 ng\u00e2n h\u00e0ng \u0111ang kh\u00f4ng ng\u1eebng \u0111\u1ed5i m\u1edbi, tr\u00ed tu\u1ec7 nh\u00e2n t\u1ea1o (AI) \u0111\u00e3 tr\u1edf th\u00e0nh y\u1ebfu t\u1ed1 ch\u1ee7 \u0111\u1ea1o th\u00fac \u0111\u1ea9y s\u1ef1 chuy\u1ec3n m\u00ecnh c\u1ee7a c\u00e1c t\u1ed5 ch\u1ee9c \u2013 t\u1eeb chi\u1ebfn l\u01b0\u1ee3c v\u1eadn h\u00e0nh cho \u0111\u1ebfn c\u00e1ch ti\u1ebfp c\u1eadn kh\u00e1ch h\u00e0ng. Tr\u00ean to\u00e0n c\u1ea7u, \u0111\u1eb7c bi\u1ec7t l\u00e0 t\u1ea1i B\u1eafc M\u1ef9, c\u00e1c ng\u00e2n h\u00e0ng l\u1edbn \u0111ang ti\u00ean phong \u1ee9ng d\u1ee5ng AI nh\u1eb1m d\u1eabn \u0111\u1ea7u \u0111\u1ed5i m\u1edbi, ph\u00e1t tri\u1ec3n n\u0103ng l\u1ef1c nh\u00e2n s\u1ef1 v\u00e0 n\u00e2ng cao t\u00ednh minh b\u1ea1ch trong ho\u1ea1t \u0111\u1ed9ng.<\/p>\n\n\n\n

Theo b\u00e1o c\u00e1o c\u1ee7a Deloitte<\/a>, nh\u1eefng ng\u00e2n h\u00e0ng \u0111\u1ea7u t\u01b0 h\u00e0ng \u0111\u1ea7u th\u1ebf gi\u1edbi c\u00f3 th\u1ec3 n\u00e2ng cao hi\u1ec7u su\u1ea5t kinh doanh l\u00ean t\u1edbi 35% nh\u1edd AI. \u0110\u00e1ng ch\u00fa \u00fd, doanh thu tr\u00ean m\u1ed7i nh\u00e2n vi\u00ean \u0111\u01b0\u1ee3c d\u1ef1 b\u00e1o s\u1ebd t\u0103ng th\u00eam 3,5 tri\u1ec7u USD v\u00e0o n\u0103m 2026, n\u1ebfu c\u00e1c gi\u1ea3i ph\u00e1p AI \u0111\u01b0\u1ee3c tri\u1ec3n khai hi\u1ec7u qu\u1ea3.<\/p>\n\n\n\n

AI kh\u00f4ng ch\u1ec9 gi\u00fap t\u1ef1 \u0111\u1ed9ng h\u00f3a quy tr\u00ecnh v\u00e0 n\u00e2ng cao hi\u1ec7u su\u1ea5t v\u1eadn h\u00e0nh, m\u00e0 c\u00f2n \u0111\u00f3ng vai tr\u00f2 then ch\u1ed1t trong vi\u1ec7c ph\u00f2ng ch\u1ed1ng gian l\u1eadn t\u00e0i ch\u00ednh. V\u00ed d\u1ee5, ng\u00e2n h\u00e0ng J.P. Morgan Chase (JPMC) \u0111\u00e3 \u1ee9ng d\u1ee5ng AI \u0111\u1ec3 c\u1ea3i thi\u1ec7n quy tr\u00ecnh x\u00e1c th\u1ef1c thanh to\u00e1n, qua \u0111\u00f3 gi\u1ea3m \u0111\u1ebfn 20% t\u1ef7 l\u1ec7 t\u1eeb ch\u1ed1i x\u00e1c th\u1ef1c t\u00e0i kho\u1ea3n<\/a> v\u00e0 ti\u1ebft ki\u1ec7m m\u1ed9t kho\u1ea3n chi ph\u00ed \u0111\u00e1ng k\u1ec3. B\u00ean c\u1ea1nh \u0111\u00f3, ng\u00e2n s\u00e1ch to\u00e0n c\u1ea7u d\u00e0nh cho AI trong ng\u00e0nh ng\u00e2n h\u00e0ng d\u1ef1 ki\u1ebfn \u0111\u1ea1t kho\u1ea3ng 85 t\u1ef7 USD v\u00e0o n\u0103m 2030<\/a>, t\u0103ng v\u1ecdt so v\u1edbi m\u1ee9c 6 t\u1ef7 USD n\u0103m 2024. <\/p>\n\n\n\n

T\u1ea1i Vi\u1ec7t Nam, \u1ee9ng d\u1ee5ng AI trong ng\u00e2n h\u00e0ng c\u0169ng \u0111ang ph\u00e1t tri\u1ec3n nhanh ch\u00f3ng v\u00e0 ng\u00e0y c\u00e0ng \u0111a d\u1ea1ng. H\u1ea7u h\u1ebft c\u00e1c ng\u00e2n h\u00e0ng l\u1edbn \u0111\u00e3 v\u00e0 \u0111ang t\u00edch c\u1ef1c tri\u1ec3n khai c\u00e1c gi\u1ea3i ph\u00e1p nh\u01b0 chatbot, tr\u1ee3 l\u00fd \u1ea3o, \u0111\u1ecbnh danh kh\u00e1ch h\u00e0ng \u0111i\u1ec7n t\u1eed (eKYC) v\u00e0 t\u1ef1 \u0111\u1ed9ng h\u00f3a quy tr\u00ecnh b\u1eb1ng robot (RPA). Ti\u00eau bi\u1ec3u:<\/p>\n\n\n\n