<\/p>
\u00a0<\/span><\/p> Avrupa Birli\u011fi Yapay Zeka T\u00fcz\u00fc\u011f\u00fc, 1 A\u011fustos 2024 tarihi itibar\u0131yla y\u00fcr\u00fcrl\u00fc\u011fe girdi. T\u00fcz\u00fck, yapay zeka sistemlerini \u00e7e\u015fitli risk seviyelerine g\u00f6re s\u0131n\u0131fland\u0131rarak, bu sistemlerin d\u00fczenlenmesi ve denetlenmesi i\u00e7in kapsaml\u0131 y\u00fck\u00fcml\u00fcl\u00fckler getiriyor.\u00a0<\/span><\/p> \u00a0<\/span><\/p> Bu y\u00fck\u00fcml\u00fcl\u00fckler, yapay zeka sistemlerinin g\u00fcvenli, adil ve \u015feffaf bir \u015fekilde kullan\u0131lmas\u0131n\u0131 sa\u011flamay\u0131 hedeflerken, bu sistemlerin insan haklar\u0131na ve temel \u00f6zg\u00fcrl\u00fcklere zarar vermemesi i\u00e7in denetim mekanizmalar\u0131 olu\u015fturuldu.\u00a0<\/span><\/p> \u00a0<\/span><\/p> Risk seviyeleri minimum risk, s\u0131n\u0131rl\u0131 risk, y\u00fcksek risk ve kabul edilemez risk kategorilerine ayr\u0131ld\u0131.<\/span><\/p> \u00a0<\/span><\/p> Avrupa Birli\u011fi Yapay Zeka T\u00fcz\u00fc\u011f\u00fc ayr\u0131ca, GPAI (General Purpose on Artificial Intelligence) modelini de dikkate alarak uluslararas\u0131 i\u015fbirli\u011fi ve k\u00fcresel standartlarla uyumu da te\u015fvik ediyor.\u00a0<\/span><\/p> \u00a0<\/span><\/p> GPAI, genel ama\u00e7l\u0131 bir yapay zeka modeline dayal\u0131 hem do\u011frudan kullan\u0131m hem de di\u011fer yapay zeka sistemlerine entegrasyon i\u00e7in \u00e7e\u015fitli ama\u00e7lara hizmet etme kapasitesine sahip bir yapay zeka sistemidir. T\u00fcz\u00fck, y\u00fcksek riskli yapay zeka sa\u011flay\u0131c\u0131lar\u0131na oldu\u011fu gibi GPAI sa\u011flay\u0131c\u0131lar\u0131na da baz\u0131 y\u00fck\u00fcml\u00fcl\u00fckler getiriyor.\u00a0<\/span><\/p> \u00a0<\/span><\/p> T\u00fcz\u00fc\u011f\u00fcn uygulanmas\u0131, kurulu\u015flara t\u00fcz\u00fc\u011f\u00fcn h\u00fck\u00fcmlerine uymalar\u0131 i\u00e7in yeterli s\u00fcre sa\u011flamak amac\u0131yla a\u015famal\u0131 olarak ger\u00e7ekle\u015ftirilecek.\u00a0<\/span><\/p> \u00a0<\/span><\/p> Bu kapsamda:<\/span><\/p> \u00a0<\/span><\/p> * 2 \u015eubat 2025 tarihinden itibaren \u00fcye devletler, kabul edilemez risk ta\u015f\u0131yan yapay zeka sistemlerini a\u015famal\u0131 olarak kald\u0131racak.<\/span><\/p> \u00a0<\/span><\/p> * 2 A\u011fustos 2025 tarihinden itibaren GPAI i\u00e7in \u00f6ng\u00f6r\u00fclen t\u00fcm y\u00fck\u00fcml\u00fcl\u00fckler uygulanacak.<\/span><\/p> \u00a0<\/span><\/p> * 2 A\u011fustos 2026 tarihinden itibaren Ek-3\u2019te yer alan y\u00fcksek riskli yapay zeka sistemleri i\u00e7in \u00f6ng\u00f6r\u00fclen t\u00fcm y\u00fck\u00fcml\u00fcl\u00fckler uygulanacak.<\/span><\/p> \u00a0<\/span><\/p> * 2 A\u011fustos 2027 tarihinden itibaren Ek-1\u2019de yer alan y\u00fcksek riskli yapay zeka sistemleri i\u00e7in \u00f6ng\u00f6r\u00fclen t\u00fcm y\u00fck\u00fcml\u00fcl\u00fckler uygulanacak.<\/span><\/p> \u00a0<\/span><\/strong><\/p> KABUL ED\u0130LEMEZ R\u0130SK TA\u015eIYAN YAPAY ZEKA S\u0130STEMLER\u0130<\/span><\/strong><\/p> \u00a0<\/span><\/p> T\u00fcz\u00fckle birlikte kabul edilemez risk (unacceptable risk) ta\u015f\u0131yan yapay zeka sistemleri yasakland\u0131.\u00a0<\/span><\/p> \u00a0<\/span><\/strong><\/p> Y\u00dcKSEK R\u0130SK TA\u015eIYAN YAPAY ZEKA S\u0130STEMLER\u0130<\/span><\/strong><\/p> \u00a0<\/span><\/p> Y\u00fcksek risk ta\u015f\u0131yan yapay zeka sistemleri t\u00fcz\u00fc\u011f\u00fcn ana odak noktas\u0131d\u0131r.<\/span><\/p> \u00a0<\/span><\/p> T\u00fcz\u00fckle birlikte yasala\u015fan \u00e7o\u011fu y\u00fck\u00fcml\u00fcl\u00fck, y\u00fcksek riskli yapay zeka sistemlerinin sa\u011flay\u0131c\u0131lar\u0131-geli\u015ftiricileri \u00fczerine y\u00fcklendi. Sa\u011flay\u0131c\u0131lar-geli\u015ftiriciler (providers-developers) genellikle sistemin yaz\u0131l\u0131m ve donan\u0131m\u0131n\u0131 geli\u015ftiren, performans\u0131n\u0131 ve g\u00fcvenli\u011fini izleyen, kullan\u0131c\u0131 e\u011fitimlerini d\u00fczenleyen ve yasal uyumlulu\u011fu sa\u011flayan kurulu\u015flar veya bireylerdir. Sa\u011flay\u0131c\u0131lar, genellikle yapay zeka sisteminin ticari \u00fcreticisi veya sunucusu olarak tan\u0131mlanabilirken, geli\u015ftiriciler sistemin teknik y\u00f6nlerinden sorumlu olan m\u00fchendisler, programc\u0131lar ve ara\u015ft\u0131rmac\u0131lard\u0131r.<\/span><\/p> \u00a0<\/span><\/p> AB\u2019de veya \u00fc\u00e7\u00fcnc\u00fc bir \u00fclkede bulunmalar\u0131na bak\u0131lmaks\u0131z\u0131n AB\u2019de y\u00fcksek riskli yapay zeka sistemlerini pazara s\u00fcrmeyi veya kullan\u0131ma sokmay\u0131 hedefleyen sa\u011flay\u0131c\u0131lar-geli\u015ftiriciler, t\u00fcz\u00fck kapsam\u0131nda sorumludur.\u00a0<\/span><\/p> \u00a0<\/span><\/p> Ayr\u0131ca, y\u00fcksek riskli yapay zeka sistemlerinin \u00e7\u0131kt\u0131s\u0131n\u0131 AB\u2019de kullanan \u00fc\u00e7\u00fcnc\u00fc \u00fclke sa\u011flay\u0131c\u0131lar\u0131 da sorumludur.<\/span><\/p> \u00a0<\/span><\/strong><\/p> SINIRLI R\u0130SK TA\u015eIYAN YAPAY ZEKA S\u0130STEMLER\u0130<\/span><\/strong><\/p> \u00a0<\/span><\/p> S\u0131n\u0131rl\u0131 risk (limited risk) ta\u015f\u0131yan yapay zeka sistemleri, kullan\u0131m\u0131nda baz\u0131 d\u00fczenlemeler gerektiren, ancak y\u00fcksek risk kategorisine girmeyen sistemlerdir. Bu t\u00fcr sistemler, kullan\u0131c\u0131lar\u0131n dikkatini \u00e7ekme ve uygun uyar\u0131lar\u0131 sa\u011flama gibi belirli \u015feffafl\u0131k (transparency) ve izleme gereksinimlerine tabi olabilirler. T\u00fcz\u00fck i\u00e7erisinde s\u0131n\u0131rl\u0131 risk i\u00e7eren yapay zeka sistemleri hakk\u0131nda \u015feffafl\u0131k sa\u011flamak amac\u0131yla son kullan\u0131c\u0131lar\u0131n yapay zeka ile etkile\u015fimde olduklar\u0131nda bilgilendirilmeleri gerekti\u011fi vurgulan\u0131yor.<\/span><\/p> \u00a0<\/span><\/strong><\/p> M\u0130N\u0130MUM R\u0130SK TA\u015eIYAN YAPAY ZEKA S\u0130STEMLER\u0130<\/span><\/strong><\/p> \u00a0<\/span><\/p> Yapay zeka destekli video oyunlar\u0131 ve spam filtreleri gibi minimum risk ta\u015f\u0131yan yapay zeka sistemleri, kullan\u0131m\u0131 genellikle serbest b\u0131rak\u0131lan ve d\u00fczenlemelerin en az oldu\u011fu kategorideki sistemlerdir. Bu sistemler, belirli k\u0131s\u0131tlamalar olmadan kullan\u0131labilir, ancak yasal uyumluluk, \u015feffafl\u0131k ve g\u00fcvenlik standartlar\u0131na uygunluk gibi temel y\u00fck\u00fcml\u00fcl\u00fckler ge\u00e7erlidir.<\/span><\/p>","slug":"-ab-yapay-zeka-tuzugu-ne-yukumlulukler-getiriyor-","tags":"AB,AvrupaBirli\u011fi,hukuk,YapayZeka,t\u00fcz\u00fck,risk,zeka","meta_title":"AB yapay zeka t\u00fcz\u00fc\u011f\u00fc ne y\u00fck\u00fcml\u00fcl\u00fckler getiriyor?","meta_description":"AV. MUHAMMET AKSAN","meta_keywords":"AB Avrupa Birli\u011fi hukuk Yapay Zeka t\u00fcz\u00fck risk zeka","news_cover_min":"\/front\/uploads\/blog\/thumbnail\/1726002000f1fRjhOHYLSfGOh.webp","news_cover":null,"news_video_min":null,"news_video":null,"view_count":8676,"cropped_1200x675":null,"user":{"id":29,"name":"MUHAMMET","surname":"AKSAN","email":"av-muhammet-aksan@gmail.com","slug":"av-muhammet-aksan","avatar":"\/front\/uploads\/avatar\/1733691600IewiFiiT1aT8ZK3.webp","status":1,"role":1,"email_verified_at":null,"orde":null,"created_at":"2022-10-24T10:51:30.000000Z","updated_at":"2024-12-18T09:45:38.000000Z","seo_title":null,"seo_description":null},"translations":[{"id":55239,"is_featured":0,"is_amp":0,"is_ads":0,"ads_link":null,"post_id":55113,"locale":"tr","category_id":73,"title":"AB yapay zeka t\u00fcz\u00fc\u011f\u00fc ne y\u00fck\u00fcml\u00fcl\u00fckler getiriyor?","home_title":null,"sub_title":null,"slider_title":null,"slider_title_2":null,"slider_spot_title":null,"slider_spot_title_2":null,"subtitleuse":0,"description":"AV. MUHAMMET AKSAN","content":" \u00a0<\/span><\/p> Avrupa Birli\u011fi Yapay Zeka T\u00fcz\u00fc\u011f\u00fc, 1 A\u011fustos 2024 tarihi itibar\u0131yla y\u00fcr\u00fcrl\u00fc\u011fe girdi. T\u00fcz\u00fck, yapay zeka sistemlerini \u00e7e\u015fitli risk seviyelerine g\u00f6re s\u0131n\u0131fland\u0131rarak, bu sistemlerin d\u00fczenlenmesi ve denetlenmesi i\u00e7in kapsaml\u0131 y\u00fck\u00fcml\u00fcl\u00fckler getiriyor.\u00a0<\/span><\/p> \u00a0<\/span><\/p> Bu y\u00fck\u00fcml\u00fcl\u00fckler, yapay zeka sistemlerinin g\u00fcvenli, adil ve \u015feffaf bir \u015fekilde kullan\u0131lmas\u0131n\u0131 sa\u011flamay\u0131 hedeflerken, bu sistemlerin insan haklar\u0131na ve temel \u00f6zg\u00fcrl\u00fcklere zarar vermemesi i\u00e7in denetim mekanizmalar\u0131 olu\u015fturuldu.\u00a0<\/span><\/p> \u00a0<\/span><\/p> Risk seviyeleri minimum risk, s\u0131n\u0131rl\u0131 risk, y\u00fcksek risk ve kabul edilemez risk kategorilerine ayr\u0131ld\u0131.<\/span><\/p> \u00a0<\/span><\/p> Avrupa Birli\u011fi Yapay Zeka T\u00fcz\u00fc\u011f\u00fc ayr\u0131ca, GPAI (General Purpose on Artificial Intelligence) modelini de dikkate alarak uluslararas\u0131 i\u015fbirli\u011fi ve k\u00fcresel standartlarla uyumu da te\u015fvik ediyor.\u00a0<\/span><\/p> \u00a0<\/span><\/p> GPAI, genel ama\u00e7l\u0131 bir yapay zeka modeline dayal\u0131 hem do\u011frudan kullan\u0131m hem de di\u011fer yapay zeka sistemlerine entegrasyon i\u00e7in \u00e7e\u015fitli ama\u00e7lara hizmet etme kapasitesine sahip bir yapay zeka sistemidir. T\u00fcz\u00fck, y\u00fcksek riskli yapay zeka sa\u011flay\u0131c\u0131lar\u0131na oldu\u011fu gibi GPAI sa\u011flay\u0131c\u0131lar\u0131na da baz\u0131 y\u00fck\u00fcml\u00fcl\u00fckler getiriyor.\u00a0<\/span><\/p> \u00a0<\/span><\/p> T\u00fcz\u00fc\u011f\u00fcn uygulanmas\u0131, kurulu\u015flara t\u00fcz\u00fc\u011f\u00fcn h\u00fck\u00fcmlerine uymalar\u0131 i\u00e7in yeterli s\u00fcre sa\u011flamak amac\u0131yla a\u015famal\u0131 olarak ger\u00e7ekle\u015ftirilecek.\u00a0<\/span><\/p> \u00a0<\/span><\/p> Bu kapsamda:<\/span><\/p> \u00a0<\/span><\/p> * 2 \u015eubat 2025 tarihinden itibaren \u00fcye devletler, kabul edilemez risk ta\u015f\u0131yan yapay zeka sistemlerini a\u015famal\u0131 olarak kald\u0131racak.<\/span><\/p> \u00a0<\/span><\/p> * 2 A\u011fustos 2025 tarihinden itibaren GPAI i\u00e7in \u00f6ng\u00f6r\u00fclen t\u00fcm y\u00fck\u00fcml\u00fcl\u00fckler uygulanacak.<\/span><\/p> \u00a0<\/span><\/p> * 2 A\u011fustos 2026 tarihinden itibaren Ek-3\u2019te yer alan y\u00fcksek riskli yapay zeka sistemleri i\u00e7in \u00f6ng\u00f6r\u00fclen t\u00fcm y\u00fck\u00fcml\u00fcl\u00fckler uygulanacak.<\/span><\/p> \u00a0<\/span><\/p> * 2 A\u011fustos 2027 tarihinden itibaren Ek-1\u2019de yer alan y\u00fcksek riskli yapay zeka sistemleri i\u00e7in \u00f6ng\u00f6r\u00fclen t\u00fcm y\u00fck\u00fcml\u00fcl\u00fckler uygulanacak.<\/span><\/p> \u00a0<\/span><\/strong><\/p> KABUL ED\u0130LEMEZ R\u0130SK TA\u015eIYAN YAPAY ZEKA S\u0130STEMLER\u0130<\/span><\/strong><\/p> \u00a0<\/span><\/p> T\u00fcz\u00fckle birlikte kabul edilemez risk (unacceptable risk) ta\u015f\u0131yan yapay zeka sistemleri yasakland\u0131.\u00a0<\/span><\/p> \u00a0<\/span><\/strong><\/p> Y\u00dcKSEK R\u0130SK TA\u015eIYAN YAPAY ZEKA S\u0130STEMLER\u0130<\/span><\/strong><\/p> \u00a0<\/span><\/p> Y\u00fcksek risk ta\u015f\u0131yan yapay zeka sistemleri t\u00fcz\u00fc\u011f\u00fcn ana odak noktas\u0131d\u0131r.<\/span><\/p> \u00a0<\/span><\/p> T\u00fcz\u00fckle birlikte yasala\u015fan \u00e7o\u011fu y\u00fck\u00fcml\u00fcl\u00fck, y\u00fcksek riskli yapay zeka sistemlerinin sa\u011flay\u0131c\u0131lar\u0131-geli\u015ftiricileri \u00fczerine y\u00fcklendi. Sa\u011flay\u0131c\u0131lar-geli\u015ftiriciler (providers-developers) genellikle sistemin yaz\u0131l\u0131m ve donan\u0131m\u0131n\u0131 geli\u015ftiren, performans\u0131n\u0131 ve g\u00fcvenli\u011fini izleyen, kullan\u0131c\u0131 e\u011fitimlerini d\u00fczenleyen ve yasal uyumlulu\u011fu sa\u011flayan kurulu\u015flar veya bireylerdir. Sa\u011flay\u0131c\u0131lar, genellikle yapay zeka sisteminin ticari \u00fcreticisi veya sunucusu olarak tan\u0131mlanabilirken, geli\u015ftiriciler sistemin teknik y\u00f6nlerinden sorumlu olan m\u00fchendisler, programc\u0131lar ve ara\u015ft\u0131rmac\u0131lard\u0131r.<\/span><\/p> \u00a0<\/span><\/p> AB\u2019de veya \u00fc\u00e7\u00fcnc\u00fc bir \u00fclkede bulunmalar\u0131na bak\u0131lmaks\u0131z\u0131n AB\u2019de y\u00fcksek riskli yapay zeka sistemlerini pazara s\u00fcrmeyi veya kullan\u0131ma sokmay\u0131 hedefleyen sa\u011flay\u0131c\u0131lar-geli\u015ftiriciler, t\u00fcz\u00fck kapsam\u0131nda sorumludur.\u00a0<\/span><\/p> \u00a0<\/span><\/p> Ayr\u0131ca, y\u00fcksek riskli yapay zeka sistemlerinin \u00e7\u0131kt\u0131s\u0131n\u0131 AB\u2019de kullanan \u00fc\u00e7\u00fcnc\u00fc \u00fclke sa\u011flay\u0131c\u0131lar\u0131 da sorumludur.<\/span><\/p> \u00a0<\/span><\/strong><\/p> SINIRLI R\u0130SK TA\u015eIYAN YAPAY ZEKA S\u0130STEMLER\u0130<\/span><\/strong><\/p> \u00a0<\/span><\/p> S\u0131n\u0131rl\u0131 risk (limited risk) ta\u015f\u0131yan yapay zeka sistemleri, kullan\u0131m\u0131nda baz\u0131 d\u00fczenlemeler gerektiren, ancak y\u00fcksek risk kategorisine girmeyen sistemlerdir. Bu t\u00fcr sistemler, kullan\u0131c\u0131lar\u0131n dikkatini \u00e7ekme ve uygun uyar\u0131lar\u0131 sa\u011flama gibi belirli \u015feffafl\u0131k (transparency) ve izleme gereksinimlerine tabi olabilirler. T\u00fcz\u00fck i\u00e7erisinde s\u0131n\u0131rl\u0131 risk i\u00e7eren yapay zeka sistemleri hakk\u0131nda \u015feffafl\u0131k sa\u011flamak amac\u0131yla son kullan\u0131c\u0131lar\u0131n yapay zeka ile etkile\u015fimde olduklar\u0131nda bilgilendirilmeleri gerekti\u011fi vurgulan\u0131yor.<\/span><\/p> \u00a0<\/span><\/strong><\/p> M\u0130N\u0130MUM R\u0130SK TA\u015eIYAN YAPAY ZEKA S\u0130STEMLER\u0130<\/span><\/strong><\/p> \u00a0<\/span><\/p> Yapay zeka destekli video oyunlar\u0131 ve spam filtreleri gibi minimum risk ta\u015f\u0131yan yapay zeka sistemleri, kullan\u0131m\u0131 genellikle serbest b\u0131rak\u0131lan ve d\u00fczenlemelerin en az oldu\u011fu kategorideki sistemlerdir. Bu sistemler, belirli k\u0131s\u0131tlamalar olmadan kullan\u0131labilir, ancak yasal uyumluluk, \u015feffafl\u0131k ve g\u00fcvenlik standartlar\u0131na uygunluk gibi temel y\u00fck\u00fcml\u00fcl\u00fckler ge\u00e7erlidir.<\/span><\/p>","slug":"-ab-yapay-zeka-tuzugu-ne-yukumlulukler-getiriyor-","orjinalimage":"\/front\/uploads\/blog\/thumbnail\/1726002000f1fRjhOHYLSfGOh.webp","news_cover_min":"\/front\/uploads\/blog\/thumbnail\/1726002000f1fRjhOHYLSfGOh.webp","news_cover":null,"news_video_min":null,"news_video":null,"cropped_638x552":null,"cropped_310x208":null,"cropped_416x247":null,"cropped_197x247":null,"cropped_416x600":null,"cropped_1200x675":null,"tags":"AB,AvrupaBirli\u011fi,hukuk,YapayZeka,t\u00fcz\u00fck,risk,zeka","meta_title":"AB yapay zeka t\u00fcz\u00fc\u011f\u00fc ne y\u00fck\u00fcml\u00fcl\u00fckler getiriyor?","meta_description":"AV. MUHAMMET AKSAN","meta_keywords":"AB Avrupa Birli\u011fi hukuk Yapay Zeka t\u00fcz\u00fck risk zeka","view_count":8676,"yt":0,"ytid":"","ytimage":null,"imgdate":"2000-01-01 00:00:00","cuff_cover":null,"cropped_358x214":null,"cropped_842x474":null}]}}
<\/p>