D\u00edky sv\u00e9mu r\u016fstu se spole\u010dnost Salesforce stala tak\u00e9 velk\u00fdm ter\u010dem kr\u00e1de\u017e\u00ed dat.<\/p>\n\n\n\n
Vzhledem k tomu, \u017ee je v\u0161e centralizov\u00e1no, m\u016f\u017ee jedin\u00e9 naru\u0161en\u00ed odhalit miliony z\u00e1znam\u016f o z\u00e1kazn\u00edc\u00edch z mnoha r\u016fzn\u00fdch spole\u010dnost\u00ed.<\/p>\n\n\n\n
Jak uv\u00e1d\u00ed Tim West, vedouc\u00ed odd\u011blen\u00ed Threat Intelligence spole\u010dnosti WithSecure:<\/p>\n\n\n\n
\"Hackersk\u00e9 skupiny jako Scattered Spider pou\u017e\u00edvaj\u00ed soci\u00e1ln\u00ed in\u017een\u00fdrstv\u00ed k z\u00edsk\u00e1n\u00ed p\u0159\u00edstupu do prost\u0159ed\u00ed SaaS. Jejich \u00fatoky mohou vypadat technicky jednodu\u0161e, ale to neznamen\u00e1, \u017ee jsou m\u00e9n\u011b nebezpe\u010dn\u00e9.\"<\/p>\n\n\n
![\"\"](\"https:\/\/blog.truthscan.com\/wp-content\/uploads\/2025\/10\/Deepfake-target-salesforce-1024x344.jpg\" "\\"\\"")
<\/figure>\n<\/div>\n\n\nPodle nov\u00e9ho pr\u016fzkumu spole\u010dnosti WithSecure se v prvn\u00edm \u010dtvrtlet\u00ed roku 2025 prudce zv\u00fd\u0161ila \u0161kodliv\u00e1 aktivita v prost\u0159ed\u00edch Salesforce, p\u0159i\u010dem\u017e po\u010det detekc\u00ed se ve srovn\u00e1n\u00ed s koncem roku 2024 zv\u00fd\u0161il dvacetin\u00e1sobn\u011b.<\/p>\n\n\n\n
Jak hlubok\u00e9 podvrhy p\u0159isp\u011bly k naru\u0161en\u00ed bezpe\u010dnosti slu\u017eby Salesforce<\/strong><\/h2>\n\n\n\nNed\u00e1vn\u00e9 \u00fatoky jsou obzvl\u00e1\u0161t\u011b nebezpe\u010dn\u00e9 proto, \u017ee se technologie deepfake zm\u011bnila ze specializovan\u00e9ho n\u00e1stroje na zbra\u0148, kterou m\u016f\u017ee pou\u017e\u00edt kdokoli.<\/p>\n\n\n\n
Na rozd\u00edl od tradi\u010dn\u00edch naru\u0161en\u00ed dat, p\u0159i kter\u00fdch se kyberzlo\u010dinci vloup\u00e1vaj\u00ed p\u0159\u00edmo do datab\u00e1z\u00ed, nyn\u00ed vyu\u017e\u00edvaj\u00ed soci\u00e1ln\u00ed in\u017een\u00fdrstv\u00ed zalo\u017een\u00e9 na hlasu neboli \"vishing\". S rozvojem deepfakes a klonov\u00e1n\u00ed hlasu pomoc\u00ed um\u011bl\u00e9 inteligence je mnohem t\u011b\u017e\u0161\u00ed tyto \u00fatoky odhalit.<\/p>\n\n\n
\n![\"\"](\"https:\/\/blog.truthscan.com\/wp-content\/uploads\/2025\/10\/Deepfake-creation-economics-1024x360.jpg\" "\\"\\"")
<\/figure>\n<\/div>\n\n\nKampa\u0148 ShinyHunters proti z\u00e1kazn\u00edk\u016fm spole\u010dnosti Salesforce se \u0159\u00edd\u00ed \u00fa\u010dinn\u00fdm n\u00e1vodem, kter\u00fd kombinuje tradi\u010dn\u00ed soci\u00e1ln\u00ed in\u017een\u00fdrstv\u00ed se \u0161pi\u010dkov\u00fdm podvodem s um\u011blou inteligenc\u00ed:<\/p>\n\n\n\n
F\u00e1ze 1: Sb\u011br hlasov\u00fdch informac\u00ed<\/strong><\/h3>\n\n\n\n\u00dato\u010dn\u00edci za\u010d\u00ednaj\u00ed z\u00edsk\u00e1v\u00e1n\u00edm zvukov\u00fdch vzork\u016f z ve\u0159ejn\u00fdch zdroj\u016f, prezentac\u00ed vedouc\u00edch pracovn\u00edk\u016f, konferen\u010dn\u00edch hovor\u016f, firemn\u00edch vide\u00ed nebo p\u0159\u00edsp\u011bvk\u016f na soci\u00e1ln\u00edch s\u00edt\u00edch. <\/p>\n\n\n\n
Sta\u010d\u00ed jim 20-30 sekund \u010dist\u00e9ho zvuku a mohou vytvo\u0159it p\u0159esv\u011bd\u010div\u00e9 hlasov\u00e9 klony.<\/p>\n\n\n\n
F\u00e1ze 2: Vishingov\u00e9 vol\u00e1n\u00ed Deepfake<\/strong><\/h3>\n\n\n\nB\u011bhem vishingov\u00e9ho hovoru \u00fato\u010dn\u00edk p\u0159esv\u011bd\u010d\u00ed ob\u011b\u0165, aby p\u0159e\u0161la na str\u00e1nku nastaven\u00ed p\u0159ipojen\u00e9 aplikace Salesforce a schv\u00e1lila fale\u0161nou verzi aplikace Data Loader, kter\u00e1 je maskov\u00e1na m\u00edrn\u011b pozm\u011bn\u011bnou zna\u010dkou.<\/p>\n\n\n\n
T\u00edm ob\u011b\u0165 nev\u011bdomky umo\u017en\u00ed \u00fato\u010dn\u00edk\u016fm ukr\u00e1st citliv\u00e1 data ze Salesforce.<\/p>\n\n\n\n
Jejich propracovanost je pozoruhodn\u00e1. V n\u011bkter\u00fdch p\u0159\u00edpadech \u00fato\u010dn\u00edci pou\u017eili deepfake audio, aby se vyd\u00e1vali za zam\u011bstnance a p\u0159esv\u011bd\u010dili pracovn\u00edky helpdesku k autorizaci neopr\u00e1vn\u011bn\u00e9ho p\u0159\u00edstupu. <\/p>\n\n\n\n
To p\u0159edstavuje v\u00fdznamn\u00fd v\u00fdvoj od tradi\u010dn\u00edho hlasov\u00e9ho phishingu, kdy se \u00fato\u010dn\u00edci jednodu\u0161e vyd\u00e1vali za autority, k vyd\u00e1v\u00e1n\u00ed se za konkr\u00e9tn\u00ed osoby s vyu\u017eit\u00edm um\u011bl\u00e9 inteligence.<\/p>\n\n\n\n
F\u00e1ze 3: Vyu\u017eit\u00ed protokolu OAuth<\/strong><\/h3>\n\n\n\nJakmile je \u0161kodliv\u00e1 aplikace autorizov\u00e1na, \u00fato\u010dn\u00edci zcela obejdou v\u00edcefaktorov\u00e9 ov\u011b\u0159ov\u00e1n\u00ed. <\/p>\n\n\n\n
Po schv\u00e1len\u00ed fale\u0161n\u00e9 aplikace z\u00edskaj\u00ed \u00fato\u010dn\u00edci dlouhodob\u00e9 tokeny OAuth, kter\u00e9 jim umo\u017en\u00ed obej\u00edt v\u00edcefaktorov\u00e9 ov\u011b\u0159ov\u00e1n\u00ed a pracovat bez spu\u0161t\u011bn\u00ed b\u011b\u017en\u00fdch bezpe\u010dnostn\u00edch upozorn\u011bn\u00ed.<\/p>\n\n\n\n
F\u00e1ze 4: Tich\u00e1 extrakce dat<\/strong><\/h3>\n\n\n\nSkupina Threat Intelligence Group spole\u010dnosti Google varovala, \u017ee \u00fato\u010dn\u00edk pou\u017eil n\u00e1stroj Python k automatizaci procesu kr\u00e1de\u017ee dat pro ka\u017edou c\u00edlovou organizaci, p\u0159i\u010dem\u017e v\u00fdzkumn\u00edci v\u011bd\u00ed o v\u00edce ne\u017e 700 potenci\u00e1ln\u011b zasa\u017een\u00fdch organizac\u00edch.<\/p>\n\n\n\n
Pro\u010d jsme zraniteln\u00ed v\u016f\u010di hlas\u016fm AI<\/strong><\/h2>\n\n\n\n\u00dasp\u011bch t\u011bchto \u00fatok\u016f vyu\u017e\u00edv\u00e1 z\u00e1kladn\u00ed lidskou vlastnost: na\u0161i tendenci v\u011b\u0159it tomu, co sly\u0161\u00edme. <\/p>\n\n\n\n
Podle jedn\u00e9 z ned\u00e1vn\u00fdch celosv\u011btov\u00fdch studi\u00ed 70 % lid\u00ed tvrd\u00ed, \u017ee si nejsou jisti, zda dok\u00e1\u017eou rozpoznat skute\u010dn\u00fd hlas od klonovan\u00e9ho.<\/p>\n\n\n\n
Tato zranitelnost se je\u0161t\u011b zvy\u0161uje ve firemn\u00edm prost\u0159ed\u00ed, kde jsou pracovn\u00edci odd\u011blen\u00ed pomoci vy\u0161koleni, aby byli vst\u0159\u00edcn\u00ed a ochotn\u00ed, a kde pr\u00e1ce na d\u00e1lku normalizovala interakce pouze se zvukem.<\/p>\n\n\n\n
Podle zpr\u00e1vy CrowdStrike 2025 Global Threat Report do\u0161lo mezi prvn\u00ed a druhou polovinou roku 2024 k n\u00e1r\u016fstu hlasov\u00fdch phishingov\u00fdch (vishingov\u00fdch) \u00fatok\u016f o 442%, a to d\u00edky taktice phishingu a vyd\u00e1v\u00e1n\u00ed se za podvodn\u00edky generovan\u00e9 um\u011blou inteligenc\u00ed.<\/p>\n\n\n\n
Bud\u00ed\u010dek pro $25 milion\u016f<\/strong><\/h2>\n\n\n\n![\"\"](\"https:\/\/blog.truthscan.com\/wp-content\/uploads\/2025\/10\/deepfake-threat-image-1024x683.jpg\" "\\"\\"")
<\/figure>\n<\/div>\n\n\nD\u016fsledky \u00fatok\u016f s vyu\u017eit\u00edm deepfake sahaj\u00ed daleko za hranice spole\u010dnosti Salesforce. <\/p>\n\n\n\n
Lupi\u010dsk\u00e1 loupe\u017e ve v\u00fd\u0161i $25 milion\u016f ve stroj\u00edrensk\u00e9 firm\u011b Arup na za\u010d\u00e1tku roku 2024, kdy \u00fato\u010dn\u00edci pou\u017eili um\u011blou inteligenci, aby se vyd\u00e1vali za n\u011bkolik vedouc\u00edch pracovn\u00edk\u016f ve videohovoru, uk\u00e1zala, \u017ee \u017e\u00e1dn\u00e1 organizace nen\u00ed v\u016f\u010di t\u00e9to hrozb\u011b imunn\u00ed. <\/p>\n\n\n\n
Podobn\u00e9 \u00fatoky se zam\u011b\u0159ily tak\u00e9 na vedouc\u00ed pracovn\u00edky v r\u016fzn\u00fdch odv\u011btv\u00edch, v\u010detn\u011b pokusu vyd\u00e1vat se za gener\u00e1ln\u00edho \u0159editele spole\u010dnosti Ferrari Benedetta Vignu pomoc\u00ed hlasov\u00fdch hovor\u016f s um\u011blou inteligenc\u00ed, kter\u00e9 napodobovaly jeho jihoitalsk\u00fd p\u0159\u00edzvuk.<\/p>\n\n\n\n
Tyto incidenty p\u0159edstavuj\u00ed to, co bezpe\u010dnostn\u00ed experti naz\u00fdvaj\u00ed \"podvody na \u0159editele 2.0\", tedy \u00fatoky, kter\u00e9 jdou nad r\u00e1mec prost\u00e9ho vyd\u00e1v\u00e1n\u00ed se za e-mail a vytv\u00e1\u0159ej\u00ed v\u00edcesmyslov\u00e9 podvody, kter\u00e9 mohou oklamat i zku\u0161en\u00e9 vedouc\u00ed pracovn\u00edky.<\/p>\n\n\n\n
Bezpe\u010dnost platformy vs. lidsk\u00e1 zranitelnost<\/strong><\/h2>\n\n\n\nSpole\u010dnost Salesforce rychle zd\u016fraznila, \u017ee tato naru\u0161en\u00ed nep\u0159edstavuj\u00ed zranitelnost jej\u00ed platformy. <\/p>\n\n\n\n
Spole\u010dnost Salesforce kampa\u0148 UNC6040 p\u0159iznala v b\u0159eznu 2025 a varovala, \u017ee \u00fato\u010dn\u00edci se vyd\u00e1vaj\u00ed za IT podporu a sna\u017e\u00ed se p\u0159im\u011bt zam\u011bstnance, aby jim p\u0159edali p\u0159ihla\u0161ovac\u00ed \u00fadaje nebo schv\u00e1lili \u0161kodliv\u00e9 p\u0159ipojen\u00e9 aplikace. <\/p>\n\n\n\n
Spole\u010dnost zd\u016fraznila, \u017ee tyto incidenty se net\u00fdkaly \u017e\u00e1dn\u00fdch zranitelnost\u00ed v jej\u00ed platform\u011b ani z nich nevypl\u00fdvaly.<\/p>\n\n\n\n
To ukazuje hlavn\u00ed v\u00fdzvu pro v\u0161echny poskytovatele SaaS: vy\u0159e\u0161it, jak zabr\u00e1nit \u00fatok\u016fm, kter\u00e9 hraj\u00ed na lidskou d\u016fv\u011bru m\u00edsto technick\u00fdch chyb.<\/p>\n\n\n\n
Spole\u010dnost Salesforce zavedla n\u011bkolik obrann\u00fdch opat\u0159en\u00ed:<\/p>\n\n\n\n
\n- Zpevn\u011bn\u00ed p\u0159ipojen\u00fdch aplikac\u00ed: Automatick\u00e9 zak\u00e1z\u00e1n\u00ed nenainstalovan\u00fdch p\u0159ipojen\u00fdch aplikac\u00ed pro nov\u00e9 u\u017eivatele<\/li>\n\n\n\n
- Omezen\u00ed toku OAuth: Zak\u00e1z\u00e1n\u00ed p\u0159ipojen\u00ed z\u00edskan\u00fdch pomoc\u00ed ur\u010dit\u00fdch autoriza\u010dn\u00edch proces\u016f<\/li>\n\n\n\n
- Roz\u0161\u00ed\u0159en\u00e9 monitorov\u00e1n\u00ed: Vylep\u0161en\u00e1 detekce podez\u0159el\u00fdch vzor\u016f autorizace aplikac\u00ed<\/li>\n\n\n\n
- Vzd\u011bl\u00e1v\u00e1n\u00ed u\u017eivatel\u016f: Zve\u0159ejn\u011bn\u00ed pokyn\u016f k rozpozn\u00e1v\u00e1n\u00ed pokus\u016f o soci\u00e1ln\u00ed in\u017een\u00fdrstv\u00ed<\/li>\n<\/ul>\n\n\n\n
V srpnu 2025 spole\u010dnost Salesforce uzav\u0159ela v\u0161echny integrace s technologiemi Salesloft, v\u010detn\u011b aplikace Drift, pot\u00e9, co zjistila, \u017ee p\u0159i souvisej\u00edc\u00edch \u00fatoc\u00edch byly odcizeny tokeny OAuth. <\/p>\n\n\n\n
Bezpe\u010dnostn\u00ed experti nyn\u00ed varuj\u00ed, \u017ee t\u00fdmy mus\u00ed naslouchat jemn\u00fdm n\u00e1znak\u016fm, jako jsou zvl\u00e1\u0161tn\u00ed pauzy, \u0161um v pozad\u00ed nebo zvukov\u00e9 z\u00e1vady, kter\u00e9 mohou odhalit, \u017ee hlas byl vytvo\u0159en um\u011blou inteligenc\u00ed.<\/p>\n\n\n\n
Reakce podniku: Nad r\u00e1mec technologick\u00fdch \u0159e\u0161en\u00ed<\/strong><\/h2>\n\n\n\nOrganizace si za\u010d\u00ednaj\u00ed uv\u011bdomovat, \u017ee obrana proti \u00fatok\u016fm s vyu\u017eit\u00edm deepfake vy\u017eaduje v\u00edce ne\u017e jen technologick\u00e1 \u0159e\u0161en\u00ed; vy\u017eaduje z\u00e1sadn\u00ed p\u0159ehodnocen\u00ed proces\u016f d\u016fv\u011bryhodnosti a ov\u011b\u0159ov\u00e1n\u00ed.<\/p>\n\n\n\n
Komunikace s nulovou d\u016fv\u011brou<\/strong><\/h3>\n\n\n\nRealityCheck od spole\u010dnosti Beyond Identity poskytuje ka\u017ed\u00e9mu \u00fa\u010dastn\u00edkovi viditeln\u00fd, ov\u011b\u0159en\u00fd identifika\u010dn\u00ed odznak, kter\u00fd je podpo\u0159en kryptografick\u00fdm ov\u011b\u0159ov\u00e1n\u00edm za\u0159\u00edzen\u00ed a pr\u016fb\u011b\u017enou kontrolou rizik, v sou\u010dasn\u00e9 dob\u011b dostupn\u00fd pro Zoom a Microsoft Teams. <\/p>\n\n\n\n
Takov\u00e1 \u0159e\u0161en\u00ed p\u0159edstavuj\u00ed posun sm\u011brem k model\u016fm \"nikdy ned\u016fv\u011b\u0159uj, v\u017edy ov\u011b\u0159uj\" pro podnikovou komunikaci.<\/p>\n\n\n\n
Roz\u0161\u00ed\u0159en\u00e9 \u0161kolic\u00ed programy<\/strong><\/h3>\n\n\n\nOrganizace, kter\u00e9 pou\u017e\u00edvaj\u00ed platformu pro simulaci deepfake od spole\u010dnosti Resemble AI, hl\u00e1s\u00ed a\u017e o 90% ni\u017e\u0161\u00ed po\u010det \u00fasp\u011b\u0161n\u00fdch \u00fatok\u016f po zaveden\u00ed platformy, kter\u00e1 vyu\u017e\u00edv\u00e1 hyperrealistick\u00e9 simulace k ilustraci toho, jak se deepfake \u00fatoky odehr\u00e1vaj\u00ed v re\u00e1ln\u00e9m sv\u011bt\u011b.<\/p>\n\n\n\n
V\u00edcekan\u00e1lov\u00e9 ov\u011b\u0159ov\u00e1n\u00ed<\/strong><\/h3>\n\n\n\nP\u0159edn\u00ed organizace zav\u00e1d\u011bj\u00ed protokoly, kter\u00e9 vy\u017eaduj\u00ed ov\u011b\u0159en\u00ed v\u00edce kan\u00e1ly pro v\u0161echny vysoce rizikov\u00e9 po\u017eadavky, bez ohledu na to, jak autentick\u00e1 se zd\u00e1 b\u00fdt p\u016fvodn\u00ed komunikace.<\/p>\n\n\n\n
Kdy\u017e u\u017e hlas nen\u00ed pravda<\/strong><\/h2>\n\n\n\nHrozba deepfake pro Salesforce a dal\u0161\u00ed podnikov\u00e9 syst\u00e9my p\u0159edstavuje zlomov\u00fd bod v historii kybernetick\u00e9 bezpe\u010dnosti. <\/p>\n\n\n\n
Poprv\u00e9 se \u00fato\u010dn\u00edci mohou nejen vyd\u00e1vat za autority, ale mohou za n\u011b i mluvit, vypadat jako oni a p\u0159esv\u011bd\u010dit i vy\u0161kolen\u00e9 bezpe\u010dnostn\u00ed profesion\u00e1ly, aby podnikli kroky, kter\u00e9 ohro\u017euj\u00ed bezpe\u010dnost organizace.<\/p>\n\n\n\n
V dob\u011b podvod\u016f generovan\u00fdch um\u011blou inteligenc\u00ed je t\u0159eba si d\u016fv\u011bru zaslou\u017eit ov\u011b\u0159en\u00edm, nikoliv p\u0159edpokl\u00e1dat, \u017ee je d\u016fv\u011bryhodn\u00e1.<\/p>\n\n\n\n
Spole\u010dnosti, kter\u00e9 si to uv\u011bdomuj\u00ed a investuj\u00ed do spr\u00e1vn\u00fdch n\u00e1stroj\u016f, proces\u016f a kultury, budou m\u00edt nejlep\u0161\u00ed pozici pro ochranu bezpe\u010dnosti i d\u016fv\u011bry ve st\u00e1le nejist\u011bj\u0161\u00edm digit\u00e1ln\u00edm sv\u011bt\u011b.<\/p>\n\n\n\n
V t\u00e9to nov\u00e9 realit\u011b nemus\u00ed b\u00fdt hlas na druh\u00e9m konci linky t\u00edm, k\u00fdm se zd\u00e1.<\/p>\n\n\n\n
V dob\u011b deepfakes je neust\u00e1l\u00e1 ostra\u017eitost cenou za bezpe\u010dnost.<\/p>","protected":false},"excerpt":{"rendered":"
V \u010dervnu 2025 obdr\u017eel zam\u011bstnanec spole\u010dnosti Google zd\u00e1nliv\u011b rutinn\u00ed telefon\u00e1t od [...]<\/p>","protected":false},"author":15,"featured_media":5339,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"rank_math_lock_modified_date":false,"_themeisle_gutenberg_block_has_review":false,"footnotes":""},"categories":[31],"tags":[],"class_list":["post-5335","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-helpful-ai-content-tips"],"_links":{"self":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/users\/15"}],"replies":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/comments?post=5335"}],"version-history":[{"count":3,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335\/revisions"}],"predecessor-version":[{"id":5343,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335\/revisions\/5343"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/media\/5339"}],"wp:attachment":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/media?parent=5335"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/categories?post=5335"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/tags?post=5335"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
<\/figure>\n<\/div>\n\n\nKampa\u0148 ShinyHunters proti z\u00e1kazn\u00edk\u016fm spole\u010dnosti Salesforce se \u0159\u00edd\u00ed \u00fa\u010dinn\u00fdm n\u00e1vodem, kter\u00fd kombinuje tradi\u010dn\u00ed soci\u00e1ln\u00ed in\u017een\u00fdrstv\u00ed se \u0161pi\u010dkov\u00fdm podvodem s um\u011blou inteligenc\u00ed:<\/p>\n\n\n\n
F\u00e1ze 1: Sb\u011br hlasov\u00fdch informac\u00ed<\/strong><\/h3>\n\n\n\n\u00dato\u010dn\u00edci za\u010d\u00ednaj\u00ed z\u00edsk\u00e1v\u00e1n\u00edm zvukov\u00fdch vzork\u016f z ve\u0159ejn\u00fdch zdroj\u016f, prezentac\u00ed vedouc\u00edch pracovn\u00edk\u016f, konferen\u010dn\u00edch hovor\u016f, firemn\u00edch vide\u00ed nebo p\u0159\u00edsp\u011bvk\u016f na soci\u00e1ln\u00edch s\u00edt\u00edch. <\/p>\n\n\n\n
Sta\u010d\u00ed jim 20-30 sekund \u010dist\u00e9ho zvuku a mohou vytvo\u0159it p\u0159esv\u011bd\u010div\u00e9 hlasov\u00e9 klony.<\/p>\n\n\n\n
F\u00e1ze 2: Vishingov\u00e9 vol\u00e1n\u00ed Deepfake<\/strong><\/h3>\n\n\n\nB\u011bhem vishingov\u00e9ho hovoru \u00fato\u010dn\u00edk p\u0159esv\u011bd\u010d\u00ed ob\u011b\u0165, aby p\u0159e\u0161la na str\u00e1nku nastaven\u00ed p\u0159ipojen\u00e9 aplikace Salesforce a schv\u00e1lila fale\u0161nou verzi aplikace Data Loader, kter\u00e1 je maskov\u00e1na m\u00edrn\u011b pozm\u011bn\u011bnou zna\u010dkou.<\/p>\n\n\n\n
T\u00edm ob\u011b\u0165 nev\u011bdomky umo\u017en\u00ed \u00fato\u010dn\u00edk\u016fm ukr\u00e1st citliv\u00e1 data ze Salesforce.<\/p>\n\n\n\n
Jejich propracovanost je pozoruhodn\u00e1. V n\u011bkter\u00fdch p\u0159\u00edpadech \u00fato\u010dn\u00edci pou\u017eili deepfake audio, aby se vyd\u00e1vali za zam\u011bstnance a p\u0159esv\u011bd\u010dili pracovn\u00edky helpdesku k autorizaci neopr\u00e1vn\u011bn\u00e9ho p\u0159\u00edstupu. <\/p>\n\n\n\n
To p\u0159edstavuje v\u00fdznamn\u00fd v\u00fdvoj od tradi\u010dn\u00edho hlasov\u00e9ho phishingu, kdy se \u00fato\u010dn\u00edci jednodu\u0161e vyd\u00e1vali za autority, k vyd\u00e1v\u00e1n\u00ed se za konkr\u00e9tn\u00ed osoby s vyu\u017eit\u00edm um\u011bl\u00e9 inteligence.<\/p>\n\n\n\n
F\u00e1ze 3: Vyu\u017eit\u00ed protokolu OAuth<\/strong><\/h3>\n\n\n\nJakmile je \u0161kodliv\u00e1 aplikace autorizov\u00e1na, \u00fato\u010dn\u00edci zcela obejdou v\u00edcefaktorov\u00e9 ov\u011b\u0159ov\u00e1n\u00ed. <\/p>\n\n\n\n
Po schv\u00e1len\u00ed fale\u0161n\u00e9 aplikace z\u00edskaj\u00ed \u00fato\u010dn\u00edci dlouhodob\u00e9 tokeny OAuth, kter\u00e9 jim umo\u017en\u00ed obej\u00edt v\u00edcefaktorov\u00e9 ov\u011b\u0159ov\u00e1n\u00ed a pracovat bez spu\u0161t\u011bn\u00ed b\u011b\u017en\u00fdch bezpe\u010dnostn\u00edch upozorn\u011bn\u00ed.<\/p>\n\n\n\n
F\u00e1ze 4: Tich\u00e1 extrakce dat<\/strong><\/h3>\n\n\n\nSkupina Threat Intelligence Group spole\u010dnosti Google varovala, \u017ee \u00fato\u010dn\u00edk pou\u017eil n\u00e1stroj Python k automatizaci procesu kr\u00e1de\u017ee dat pro ka\u017edou c\u00edlovou organizaci, p\u0159i\u010dem\u017e v\u00fdzkumn\u00edci v\u011bd\u00ed o v\u00edce ne\u017e 700 potenci\u00e1ln\u011b zasa\u017een\u00fdch organizac\u00edch.<\/p>\n\n\n\n
Pro\u010d jsme zraniteln\u00ed v\u016f\u010di hlas\u016fm AI<\/strong><\/h2>\n\n\n\n\u00dasp\u011bch t\u011bchto \u00fatok\u016f vyu\u017e\u00edv\u00e1 z\u00e1kladn\u00ed lidskou vlastnost: na\u0161i tendenci v\u011b\u0159it tomu, co sly\u0161\u00edme. <\/p>\n\n\n\n
Podle jedn\u00e9 z ned\u00e1vn\u00fdch celosv\u011btov\u00fdch studi\u00ed 70 % lid\u00ed tvrd\u00ed, \u017ee si nejsou jisti, zda dok\u00e1\u017eou rozpoznat skute\u010dn\u00fd hlas od klonovan\u00e9ho.<\/p>\n\n\n\n
Tato zranitelnost se je\u0161t\u011b zvy\u0161uje ve firemn\u00edm prost\u0159ed\u00ed, kde jsou pracovn\u00edci odd\u011blen\u00ed pomoci vy\u0161koleni, aby byli vst\u0159\u00edcn\u00ed a ochotn\u00ed, a kde pr\u00e1ce na d\u00e1lku normalizovala interakce pouze se zvukem.<\/p>\n\n\n\n
Podle zpr\u00e1vy CrowdStrike 2025 Global Threat Report do\u0161lo mezi prvn\u00ed a druhou polovinou roku 2024 k n\u00e1r\u016fstu hlasov\u00fdch phishingov\u00fdch (vishingov\u00fdch) \u00fatok\u016f o 442%, a to d\u00edky taktice phishingu a vyd\u00e1v\u00e1n\u00ed se za podvodn\u00edky generovan\u00e9 um\u011blou inteligenc\u00ed.<\/p>\n\n\n\n
Bud\u00ed\u010dek pro $25 milion\u016f<\/strong><\/h2>\n\n\n\n<\/figure>\n<\/div>\n\n\nD\u016fsledky \u00fatok\u016f s vyu\u017eit\u00edm deepfake sahaj\u00ed daleko za hranice spole\u010dnosti Salesforce. <\/p>\n\n\n\n
Lupi\u010dsk\u00e1 loupe\u017e ve v\u00fd\u0161i $25 milion\u016f ve stroj\u00edrensk\u00e9 firm\u011b Arup na za\u010d\u00e1tku roku 2024, kdy \u00fato\u010dn\u00edci pou\u017eili um\u011blou inteligenci, aby se vyd\u00e1vali za n\u011bkolik vedouc\u00edch pracovn\u00edk\u016f ve videohovoru, uk\u00e1zala, \u017ee \u017e\u00e1dn\u00e1 organizace nen\u00ed v\u016f\u010di t\u00e9to hrozb\u011b imunn\u00ed. <\/p>\n\n\n\n
Podobn\u00e9 \u00fatoky se zam\u011b\u0159ily tak\u00e9 na vedouc\u00ed pracovn\u00edky v r\u016fzn\u00fdch odv\u011btv\u00edch, v\u010detn\u011b pokusu vyd\u00e1vat se za gener\u00e1ln\u00edho \u0159editele spole\u010dnosti Ferrari Benedetta Vignu pomoc\u00ed hlasov\u00fdch hovor\u016f s um\u011blou inteligenc\u00ed, kter\u00e9 napodobovaly jeho jihoitalsk\u00fd p\u0159\u00edzvuk.<\/p>\n\n\n\n
Tyto incidenty p\u0159edstavuj\u00ed to, co bezpe\u010dnostn\u00ed experti naz\u00fdvaj\u00ed \"podvody na \u0159editele 2.0\", tedy \u00fatoky, kter\u00e9 jdou nad r\u00e1mec prost\u00e9ho vyd\u00e1v\u00e1n\u00ed se za e-mail a vytv\u00e1\u0159ej\u00ed v\u00edcesmyslov\u00e9 podvody, kter\u00e9 mohou oklamat i zku\u0161en\u00e9 vedouc\u00ed pracovn\u00edky.<\/p>\n\n\n\n
Bezpe\u010dnost platformy vs. lidsk\u00e1 zranitelnost<\/strong><\/h2>\n\n\n\nSpole\u010dnost Salesforce rychle zd\u016fraznila, \u017ee tato naru\u0161en\u00ed nep\u0159edstavuj\u00ed zranitelnost jej\u00ed platformy. <\/p>\n\n\n\n
Spole\u010dnost Salesforce kampa\u0148 UNC6040 p\u0159iznala v b\u0159eznu 2025 a varovala, \u017ee \u00fato\u010dn\u00edci se vyd\u00e1vaj\u00ed za IT podporu a sna\u017e\u00ed se p\u0159im\u011bt zam\u011bstnance, aby jim p\u0159edali p\u0159ihla\u0161ovac\u00ed \u00fadaje nebo schv\u00e1lili \u0161kodliv\u00e9 p\u0159ipojen\u00e9 aplikace. <\/p>\n\n\n\n
Spole\u010dnost zd\u016fraznila, \u017ee tyto incidenty se net\u00fdkaly \u017e\u00e1dn\u00fdch zranitelnost\u00ed v jej\u00ed platform\u011b ani z nich nevypl\u00fdvaly.<\/p>\n\n\n\n
To ukazuje hlavn\u00ed v\u00fdzvu pro v\u0161echny poskytovatele SaaS: vy\u0159e\u0161it, jak zabr\u00e1nit \u00fatok\u016fm, kter\u00e9 hraj\u00ed na lidskou d\u016fv\u011bru m\u00edsto technick\u00fdch chyb.<\/p>\n\n\n\n
Spole\u010dnost Salesforce zavedla n\u011bkolik obrann\u00fdch opat\u0159en\u00ed:<\/p>\n\n\n\n
\n- Zpevn\u011bn\u00ed p\u0159ipojen\u00fdch aplikac\u00ed: Automatick\u00e9 zak\u00e1z\u00e1n\u00ed nenainstalovan\u00fdch p\u0159ipojen\u00fdch aplikac\u00ed pro nov\u00e9 u\u017eivatele<\/li>\n\n\n\n
- Omezen\u00ed toku OAuth: Zak\u00e1z\u00e1n\u00ed p\u0159ipojen\u00ed z\u00edskan\u00fdch pomoc\u00ed ur\u010dit\u00fdch autoriza\u010dn\u00edch proces\u016f<\/li>\n\n\n\n
- Roz\u0161\u00ed\u0159en\u00e9 monitorov\u00e1n\u00ed: Vylep\u0161en\u00e1 detekce podez\u0159el\u00fdch vzor\u016f autorizace aplikac\u00ed<\/li>\n\n\n\n
- Vzd\u011bl\u00e1v\u00e1n\u00ed u\u017eivatel\u016f: Zve\u0159ejn\u011bn\u00ed pokyn\u016f k rozpozn\u00e1v\u00e1n\u00ed pokus\u016f o soci\u00e1ln\u00ed in\u017een\u00fdrstv\u00ed<\/li>\n<\/ul>\n\n\n\n
V srpnu 2025 spole\u010dnost Salesforce uzav\u0159ela v\u0161echny integrace s technologiemi Salesloft, v\u010detn\u011b aplikace Drift, pot\u00e9, co zjistila, \u017ee p\u0159i souvisej\u00edc\u00edch \u00fatoc\u00edch byly odcizeny tokeny OAuth. <\/p>\n\n\n\n
Bezpe\u010dnostn\u00ed experti nyn\u00ed varuj\u00ed, \u017ee t\u00fdmy mus\u00ed naslouchat jemn\u00fdm n\u00e1znak\u016fm, jako jsou zvl\u00e1\u0161tn\u00ed pauzy, \u0161um v pozad\u00ed nebo zvukov\u00e9 z\u00e1vady, kter\u00e9 mohou odhalit, \u017ee hlas byl vytvo\u0159en um\u011blou inteligenc\u00ed.<\/p>\n\n\n\n
Reakce podniku: Nad r\u00e1mec technologick\u00fdch \u0159e\u0161en\u00ed<\/strong><\/h2>\n\n\n\nOrganizace si za\u010d\u00ednaj\u00ed uv\u011bdomovat, \u017ee obrana proti \u00fatok\u016fm s vyu\u017eit\u00edm deepfake vy\u017eaduje v\u00edce ne\u017e jen technologick\u00e1 \u0159e\u0161en\u00ed; vy\u017eaduje z\u00e1sadn\u00ed p\u0159ehodnocen\u00ed proces\u016f d\u016fv\u011bryhodnosti a ov\u011b\u0159ov\u00e1n\u00ed.<\/p>\n\n\n\n
Komunikace s nulovou d\u016fv\u011brou<\/strong><\/h3>\n\n\n\nRealityCheck od spole\u010dnosti Beyond Identity poskytuje ka\u017ed\u00e9mu \u00fa\u010dastn\u00edkovi viditeln\u00fd, ov\u011b\u0159en\u00fd identifika\u010dn\u00ed odznak, kter\u00fd je podpo\u0159en kryptografick\u00fdm ov\u011b\u0159ov\u00e1n\u00edm za\u0159\u00edzen\u00ed a pr\u016fb\u011b\u017enou kontrolou rizik, v sou\u010dasn\u00e9 dob\u011b dostupn\u00fd pro Zoom a Microsoft Teams. <\/p>\n\n\n\n
Takov\u00e1 \u0159e\u0161en\u00ed p\u0159edstavuj\u00ed posun sm\u011brem k model\u016fm \"nikdy ned\u016fv\u011b\u0159uj, v\u017edy ov\u011b\u0159uj\" pro podnikovou komunikaci.<\/p>\n\n\n\n
Roz\u0161\u00ed\u0159en\u00e9 \u0161kolic\u00ed programy<\/strong><\/h3>\n\n\n\nOrganizace, kter\u00e9 pou\u017e\u00edvaj\u00ed platformu pro simulaci deepfake od spole\u010dnosti Resemble AI, hl\u00e1s\u00ed a\u017e o 90% ni\u017e\u0161\u00ed po\u010det \u00fasp\u011b\u0161n\u00fdch \u00fatok\u016f po zaveden\u00ed platformy, kter\u00e1 vyu\u017e\u00edv\u00e1 hyperrealistick\u00e9 simulace k ilustraci toho, jak se deepfake \u00fatoky odehr\u00e1vaj\u00ed v re\u00e1ln\u00e9m sv\u011bt\u011b.<\/p>\n\n\n\n
V\u00edcekan\u00e1lov\u00e9 ov\u011b\u0159ov\u00e1n\u00ed<\/strong><\/h3>\n\n\n\nP\u0159edn\u00ed organizace zav\u00e1d\u011bj\u00ed protokoly, kter\u00e9 vy\u017eaduj\u00ed ov\u011b\u0159en\u00ed v\u00edce kan\u00e1ly pro v\u0161echny vysoce rizikov\u00e9 po\u017eadavky, bez ohledu na to, jak autentick\u00e1 se zd\u00e1 b\u00fdt p\u016fvodn\u00ed komunikace.<\/p>\n\n\n\n
Kdy\u017e u\u017e hlas nen\u00ed pravda<\/strong><\/h2>\n\n\n\nHrozba deepfake pro Salesforce a dal\u0161\u00ed podnikov\u00e9 syst\u00e9my p\u0159edstavuje zlomov\u00fd bod v historii kybernetick\u00e9 bezpe\u010dnosti. <\/p>\n\n\n\n
Poprv\u00e9 se \u00fato\u010dn\u00edci mohou nejen vyd\u00e1vat za autority, ale mohou za n\u011b i mluvit, vypadat jako oni a p\u0159esv\u011bd\u010dit i vy\u0161kolen\u00e9 bezpe\u010dnostn\u00ed profesion\u00e1ly, aby podnikli kroky, kter\u00e9 ohro\u017euj\u00ed bezpe\u010dnost organizace.<\/p>\n\n\n\n
V dob\u011b podvod\u016f generovan\u00fdch um\u011blou inteligenc\u00ed je t\u0159eba si d\u016fv\u011bru zaslou\u017eit ov\u011b\u0159en\u00edm, nikoliv p\u0159edpokl\u00e1dat, \u017ee je d\u016fv\u011bryhodn\u00e1.<\/p>\n\n\n\n
Spole\u010dnosti, kter\u00e9 si to uv\u011bdomuj\u00ed a investuj\u00ed do spr\u00e1vn\u00fdch n\u00e1stroj\u016f, proces\u016f a kultury, budou m\u00edt nejlep\u0161\u00ed pozici pro ochranu bezpe\u010dnosti i d\u016fv\u011bry ve st\u00e1le nejist\u011bj\u0161\u00edm digit\u00e1ln\u00edm sv\u011bt\u011b.<\/p>\n\n\n\n
V t\u00e9to nov\u00e9 realit\u011b nemus\u00ed b\u00fdt hlas na druh\u00e9m konci linky t\u00edm, k\u00fdm se zd\u00e1.<\/p>\n\n\n\n
V dob\u011b deepfakes je neust\u00e1l\u00e1 ostra\u017eitost cenou za bezpe\u010dnost.<\/p>","protected":false},"excerpt":{"rendered":"
V \u010dervnu 2025 obdr\u017eel zam\u011bstnanec spole\u010dnosti Google zd\u00e1nliv\u011b rutinn\u00ed telefon\u00e1t od [...]<\/p>","protected":false},"author":15,"featured_media":5339,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"rank_math_lock_modified_date":false,"_themeisle_gutenberg_block_has_review":false,"footnotes":""},"categories":[31],"tags":[],"class_list":["post-5335","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-helpful-ai-content-tips"],"_links":{"self":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/users\/15"}],"replies":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/comments?post=5335"}],"version-history":[{"count":3,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335\/revisions"}],"predecessor-version":[{"id":5343,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335\/revisions\/5343"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/media\/5339"}],"wp:attachment":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/media?parent=5335"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/categories?post=5335"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/tags?post=5335"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
B\u011bhem vishingov\u00e9ho hovoru \u00fato\u010dn\u00edk p\u0159esv\u011bd\u010d\u00ed ob\u011b\u0165, aby p\u0159e\u0161la na str\u00e1nku nastaven\u00ed p\u0159ipojen\u00e9 aplikace Salesforce a schv\u00e1lila fale\u0161nou verzi aplikace Data Loader, kter\u00e1 je maskov\u00e1na m\u00edrn\u011b pozm\u011bn\u011bnou zna\u010dkou.<\/p>\n\n\n\n
T\u00edm ob\u011b\u0165 nev\u011bdomky umo\u017en\u00ed \u00fato\u010dn\u00edk\u016fm ukr\u00e1st citliv\u00e1 data ze Salesforce.<\/p>\n\n\n\n
Jejich propracovanost je pozoruhodn\u00e1. V n\u011bkter\u00fdch p\u0159\u00edpadech \u00fato\u010dn\u00edci pou\u017eili deepfake audio, aby se vyd\u00e1vali za zam\u011bstnance a p\u0159esv\u011bd\u010dili pracovn\u00edky helpdesku k autorizaci neopr\u00e1vn\u011bn\u00e9ho p\u0159\u00edstupu. <\/p>\n\n\n\n
To p\u0159edstavuje v\u00fdznamn\u00fd v\u00fdvoj od tradi\u010dn\u00edho hlasov\u00e9ho phishingu, kdy se \u00fato\u010dn\u00edci jednodu\u0161e vyd\u00e1vali za autority, k vyd\u00e1v\u00e1n\u00ed se za konkr\u00e9tn\u00ed osoby s vyu\u017eit\u00edm um\u011bl\u00e9 inteligence.<\/p>\n\n\n\n
F\u00e1ze 3: Vyu\u017eit\u00ed protokolu OAuth<\/strong><\/h3>\n\n\n\nJakmile je \u0161kodliv\u00e1 aplikace autorizov\u00e1na, \u00fato\u010dn\u00edci zcela obejdou v\u00edcefaktorov\u00e9 ov\u011b\u0159ov\u00e1n\u00ed. <\/p>\n\n\n\n
Po schv\u00e1len\u00ed fale\u0161n\u00e9 aplikace z\u00edskaj\u00ed \u00fato\u010dn\u00edci dlouhodob\u00e9 tokeny OAuth, kter\u00e9 jim umo\u017en\u00ed obej\u00edt v\u00edcefaktorov\u00e9 ov\u011b\u0159ov\u00e1n\u00ed a pracovat bez spu\u0161t\u011bn\u00ed b\u011b\u017en\u00fdch bezpe\u010dnostn\u00edch upozorn\u011bn\u00ed.<\/p>\n\n\n\n
F\u00e1ze 4: Tich\u00e1 extrakce dat<\/strong><\/h3>\n\n\n\nSkupina Threat Intelligence Group spole\u010dnosti Google varovala, \u017ee \u00fato\u010dn\u00edk pou\u017eil n\u00e1stroj Python k automatizaci procesu kr\u00e1de\u017ee dat pro ka\u017edou c\u00edlovou organizaci, p\u0159i\u010dem\u017e v\u00fdzkumn\u00edci v\u011bd\u00ed o v\u00edce ne\u017e 700 potenci\u00e1ln\u011b zasa\u017een\u00fdch organizac\u00edch.<\/p>\n\n\n\n
Pro\u010d jsme zraniteln\u00ed v\u016f\u010di hlas\u016fm AI<\/strong><\/h2>\n\n\n\n\u00dasp\u011bch t\u011bchto \u00fatok\u016f vyu\u017e\u00edv\u00e1 z\u00e1kladn\u00ed lidskou vlastnost: na\u0161i tendenci v\u011b\u0159it tomu, co sly\u0161\u00edme. <\/p>\n\n\n\n
Podle jedn\u00e9 z ned\u00e1vn\u00fdch celosv\u011btov\u00fdch studi\u00ed 70 % lid\u00ed tvrd\u00ed, \u017ee si nejsou jisti, zda dok\u00e1\u017eou rozpoznat skute\u010dn\u00fd hlas od klonovan\u00e9ho.<\/p>\n\n\n\n
Tato zranitelnost se je\u0161t\u011b zvy\u0161uje ve firemn\u00edm prost\u0159ed\u00ed, kde jsou pracovn\u00edci odd\u011blen\u00ed pomoci vy\u0161koleni, aby byli vst\u0159\u00edcn\u00ed a ochotn\u00ed, a kde pr\u00e1ce na d\u00e1lku normalizovala interakce pouze se zvukem.<\/p>\n\n\n\n
Podle zpr\u00e1vy CrowdStrike 2025 Global Threat Report do\u0161lo mezi prvn\u00ed a druhou polovinou roku 2024 k n\u00e1r\u016fstu hlasov\u00fdch phishingov\u00fdch (vishingov\u00fdch) \u00fatok\u016f o 442%, a to d\u00edky taktice phishingu a vyd\u00e1v\u00e1n\u00ed se za podvodn\u00edky generovan\u00e9 um\u011blou inteligenc\u00ed.<\/p>\n\n\n\n
Bud\u00ed\u010dek pro $25 milion\u016f<\/strong><\/h2>\n\n\n\n<\/figure>\n<\/div>\n\n\nD\u016fsledky \u00fatok\u016f s vyu\u017eit\u00edm deepfake sahaj\u00ed daleko za hranice spole\u010dnosti Salesforce. <\/p>\n\n\n\n
Lupi\u010dsk\u00e1 loupe\u017e ve v\u00fd\u0161i $25 milion\u016f ve stroj\u00edrensk\u00e9 firm\u011b Arup na za\u010d\u00e1tku roku 2024, kdy \u00fato\u010dn\u00edci pou\u017eili um\u011blou inteligenci, aby se vyd\u00e1vali za n\u011bkolik vedouc\u00edch pracovn\u00edk\u016f ve videohovoru, uk\u00e1zala, \u017ee \u017e\u00e1dn\u00e1 organizace nen\u00ed v\u016f\u010di t\u00e9to hrozb\u011b imunn\u00ed. <\/p>\n\n\n\n
Podobn\u00e9 \u00fatoky se zam\u011b\u0159ily tak\u00e9 na vedouc\u00ed pracovn\u00edky v r\u016fzn\u00fdch odv\u011btv\u00edch, v\u010detn\u011b pokusu vyd\u00e1vat se za gener\u00e1ln\u00edho \u0159editele spole\u010dnosti Ferrari Benedetta Vignu pomoc\u00ed hlasov\u00fdch hovor\u016f s um\u011blou inteligenc\u00ed, kter\u00e9 napodobovaly jeho jihoitalsk\u00fd p\u0159\u00edzvuk.<\/p>\n\n\n\n
Tyto incidenty p\u0159edstavuj\u00ed to, co bezpe\u010dnostn\u00ed experti naz\u00fdvaj\u00ed \"podvody na \u0159editele 2.0\", tedy \u00fatoky, kter\u00e9 jdou nad r\u00e1mec prost\u00e9ho vyd\u00e1v\u00e1n\u00ed se za e-mail a vytv\u00e1\u0159ej\u00ed v\u00edcesmyslov\u00e9 podvody, kter\u00e9 mohou oklamat i zku\u0161en\u00e9 vedouc\u00ed pracovn\u00edky.<\/p>\n\n\n\n
Bezpe\u010dnost platformy vs. lidsk\u00e1 zranitelnost<\/strong><\/h2>\n\n\n\nSpole\u010dnost Salesforce rychle zd\u016fraznila, \u017ee tato naru\u0161en\u00ed nep\u0159edstavuj\u00ed zranitelnost jej\u00ed platformy. <\/p>\n\n\n\n
Spole\u010dnost Salesforce kampa\u0148 UNC6040 p\u0159iznala v b\u0159eznu 2025 a varovala, \u017ee \u00fato\u010dn\u00edci se vyd\u00e1vaj\u00ed za IT podporu a sna\u017e\u00ed se p\u0159im\u011bt zam\u011bstnance, aby jim p\u0159edali p\u0159ihla\u0161ovac\u00ed \u00fadaje nebo schv\u00e1lili \u0161kodliv\u00e9 p\u0159ipojen\u00e9 aplikace. <\/p>\n\n\n\n
Spole\u010dnost zd\u016fraznila, \u017ee tyto incidenty se net\u00fdkaly \u017e\u00e1dn\u00fdch zranitelnost\u00ed v jej\u00ed platform\u011b ani z nich nevypl\u00fdvaly.<\/p>\n\n\n\n
To ukazuje hlavn\u00ed v\u00fdzvu pro v\u0161echny poskytovatele SaaS: vy\u0159e\u0161it, jak zabr\u00e1nit \u00fatok\u016fm, kter\u00e9 hraj\u00ed na lidskou d\u016fv\u011bru m\u00edsto technick\u00fdch chyb.<\/p>\n\n\n\n
Spole\u010dnost Salesforce zavedla n\u011bkolik obrann\u00fdch opat\u0159en\u00ed:<\/p>\n\n\n\n
\n- Zpevn\u011bn\u00ed p\u0159ipojen\u00fdch aplikac\u00ed: Automatick\u00e9 zak\u00e1z\u00e1n\u00ed nenainstalovan\u00fdch p\u0159ipojen\u00fdch aplikac\u00ed pro nov\u00e9 u\u017eivatele<\/li>\n\n\n\n
- Omezen\u00ed toku OAuth: Zak\u00e1z\u00e1n\u00ed p\u0159ipojen\u00ed z\u00edskan\u00fdch pomoc\u00ed ur\u010dit\u00fdch autoriza\u010dn\u00edch proces\u016f<\/li>\n\n\n\n
- Roz\u0161\u00ed\u0159en\u00e9 monitorov\u00e1n\u00ed: Vylep\u0161en\u00e1 detekce podez\u0159el\u00fdch vzor\u016f autorizace aplikac\u00ed<\/li>\n\n\n\n
- Vzd\u011bl\u00e1v\u00e1n\u00ed u\u017eivatel\u016f: Zve\u0159ejn\u011bn\u00ed pokyn\u016f k rozpozn\u00e1v\u00e1n\u00ed pokus\u016f o soci\u00e1ln\u00ed in\u017een\u00fdrstv\u00ed<\/li>\n<\/ul>\n\n\n\n
V srpnu 2025 spole\u010dnost Salesforce uzav\u0159ela v\u0161echny integrace s technologiemi Salesloft, v\u010detn\u011b aplikace Drift, pot\u00e9, co zjistila, \u017ee p\u0159i souvisej\u00edc\u00edch \u00fatoc\u00edch byly odcizeny tokeny OAuth. <\/p>\n\n\n\n
Bezpe\u010dnostn\u00ed experti nyn\u00ed varuj\u00ed, \u017ee t\u00fdmy mus\u00ed naslouchat jemn\u00fdm n\u00e1znak\u016fm, jako jsou zvl\u00e1\u0161tn\u00ed pauzy, \u0161um v pozad\u00ed nebo zvukov\u00e9 z\u00e1vady, kter\u00e9 mohou odhalit, \u017ee hlas byl vytvo\u0159en um\u011blou inteligenc\u00ed.<\/p>\n\n\n\n
Reakce podniku: Nad r\u00e1mec technologick\u00fdch \u0159e\u0161en\u00ed<\/strong><\/h2>\n\n\n\nOrganizace si za\u010d\u00ednaj\u00ed uv\u011bdomovat, \u017ee obrana proti \u00fatok\u016fm s vyu\u017eit\u00edm deepfake vy\u017eaduje v\u00edce ne\u017e jen technologick\u00e1 \u0159e\u0161en\u00ed; vy\u017eaduje z\u00e1sadn\u00ed p\u0159ehodnocen\u00ed proces\u016f d\u016fv\u011bryhodnosti a ov\u011b\u0159ov\u00e1n\u00ed.<\/p>\n\n\n\n
Komunikace s nulovou d\u016fv\u011brou<\/strong><\/h3>\n\n\n\nRealityCheck od spole\u010dnosti Beyond Identity poskytuje ka\u017ed\u00e9mu \u00fa\u010dastn\u00edkovi viditeln\u00fd, ov\u011b\u0159en\u00fd identifika\u010dn\u00ed odznak, kter\u00fd je podpo\u0159en kryptografick\u00fdm ov\u011b\u0159ov\u00e1n\u00edm za\u0159\u00edzen\u00ed a pr\u016fb\u011b\u017enou kontrolou rizik, v sou\u010dasn\u00e9 dob\u011b dostupn\u00fd pro Zoom a Microsoft Teams. <\/p>\n\n\n\n
Takov\u00e1 \u0159e\u0161en\u00ed p\u0159edstavuj\u00ed posun sm\u011brem k model\u016fm \"nikdy ned\u016fv\u011b\u0159uj, v\u017edy ov\u011b\u0159uj\" pro podnikovou komunikaci.<\/p>\n\n\n\n
Roz\u0161\u00ed\u0159en\u00e9 \u0161kolic\u00ed programy<\/strong><\/h3>\n\n\n\nOrganizace, kter\u00e9 pou\u017e\u00edvaj\u00ed platformu pro simulaci deepfake od spole\u010dnosti Resemble AI, hl\u00e1s\u00ed a\u017e o 90% ni\u017e\u0161\u00ed po\u010det \u00fasp\u011b\u0161n\u00fdch \u00fatok\u016f po zaveden\u00ed platformy, kter\u00e1 vyu\u017e\u00edv\u00e1 hyperrealistick\u00e9 simulace k ilustraci toho, jak se deepfake \u00fatoky odehr\u00e1vaj\u00ed v re\u00e1ln\u00e9m sv\u011bt\u011b.<\/p>\n\n\n\n
V\u00edcekan\u00e1lov\u00e9 ov\u011b\u0159ov\u00e1n\u00ed<\/strong><\/h3>\n\n\n\nP\u0159edn\u00ed organizace zav\u00e1d\u011bj\u00ed protokoly, kter\u00e9 vy\u017eaduj\u00ed ov\u011b\u0159en\u00ed v\u00edce kan\u00e1ly pro v\u0161echny vysoce rizikov\u00e9 po\u017eadavky, bez ohledu na to, jak autentick\u00e1 se zd\u00e1 b\u00fdt p\u016fvodn\u00ed komunikace.<\/p>\n\n\n\n
Kdy\u017e u\u017e hlas nen\u00ed pravda<\/strong><\/h2>\n\n\n\nHrozba deepfake pro Salesforce a dal\u0161\u00ed podnikov\u00e9 syst\u00e9my p\u0159edstavuje zlomov\u00fd bod v historii kybernetick\u00e9 bezpe\u010dnosti. <\/p>\n\n\n\n
Poprv\u00e9 se \u00fato\u010dn\u00edci mohou nejen vyd\u00e1vat za autority, ale mohou za n\u011b i mluvit, vypadat jako oni a p\u0159esv\u011bd\u010dit i vy\u0161kolen\u00e9 bezpe\u010dnostn\u00ed profesion\u00e1ly, aby podnikli kroky, kter\u00e9 ohro\u017euj\u00ed bezpe\u010dnost organizace.<\/p>\n\n\n\n
V dob\u011b podvod\u016f generovan\u00fdch um\u011blou inteligenc\u00ed je t\u0159eba si d\u016fv\u011bru zaslou\u017eit ov\u011b\u0159en\u00edm, nikoliv p\u0159edpokl\u00e1dat, \u017ee je d\u016fv\u011bryhodn\u00e1.<\/p>\n\n\n\n
Spole\u010dnosti, kter\u00e9 si to uv\u011bdomuj\u00ed a investuj\u00ed do spr\u00e1vn\u00fdch n\u00e1stroj\u016f, proces\u016f a kultury, budou m\u00edt nejlep\u0161\u00ed pozici pro ochranu bezpe\u010dnosti i d\u016fv\u011bry ve st\u00e1le nejist\u011bj\u0161\u00edm digit\u00e1ln\u00edm sv\u011bt\u011b.<\/p>\n\n\n\n
V t\u00e9to nov\u00e9 realit\u011b nemus\u00ed b\u00fdt hlas na druh\u00e9m konci linky t\u00edm, k\u00fdm se zd\u00e1.<\/p>\n\n\n\n
V dob\u011b deepfakes je neust\u00e1l\u00e1 ostra\u017eitost cenou za bezpe\u010dnost.<\/p>","protected":false},"excerpt":{"rendered":"
V \u010dervnu 2025 obdr\u017eel zam\u011bstnanec spole\u010dnosti Google zd\u00e1nliv\u011b rutinn\u00ed telefon\u00e1t od [...]<\/p>","protected":false},"author":15,"featured_media":5339,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"rank_math_lock_modified_date":false,"_themeisle_gutenberg_block_has_review":false,"footnotes":""},"categories":[31],"tags":[],"class_list":["post-5335","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-helpful-ai-content-tips"],"_links":{"self":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/users\/15"}],"replies":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/comments?post=5335"}],"version-history":[{"count":3,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335\/revisions"}],"predecessor-version":[{"id":5343,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335\/revisions\/5343"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/media\/5339"}],"wp:attachment":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/media?parent=5335"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/categories?post=5335"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/tags?post=5335"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
Skupina Threat Intelligence Group spole\u010dnosti Google varovala, \u017ee \u00fato\u010dn\u00edk pou\u017eil n\u00e1stroj Python k automatizaci procesu kr\u00e1de\u017ee dat pro ka\u017edou c\u00edlovou organizaci, p\u0159i\u010dem\u017e v\u00fdzkumn\u00edci v\u011bd\u00ed o v\u00edce ne\u017e 700 potenci\u00e1ln\u011b zasa\u017een\u00fdch organizac\u00edch.<\/p>\n\n\n\n
Pro\u010d jsme zraniteln\u00ed v\u016f\u010di hlas\u016fm AI<\/strong><\/h2>\n\n\n\n\u00dasp\u011bch t\u011bchto \u00fatok\u016f vyu\u017e\u00edv\u00e1 z\u00e1kladn\u00ed lidskou vlastnost: na\u0161i tendenci v\u011b\u0159it tomu, co sly\u0161\u00edme. <\/p>\n\n\n\n
Podle jedn\u00e9 z ned\u00e1vn\u00fdch celosv\u011btov\u00fdch studi\u00ed 70 % lid\u00ed tvrd\u00ed, \u017ee si nejsou jisti, zda dok\u00e1\u017eou rozpoznat skute\u010dn\u00fd hlas od klonovan\u00e9ho.<\/p>\n\n\n\n
Tato zranitelnost se je\u0161t\u011b zvy\u0161uje ve firemn\u00edm prost\u0159ed\u00ed, kde jsou pracovn\u00edci odd\u011blen\u00ed pomoci vy\u0161koleni, aby byli vst\u0159\u00edcn\u00ed a ochotn\u00ed, a kde pr\u00e1ce na d\u00e1lku normalizovala interakce pouze se zvukem.<\/p>\n\n\n\n
Podle zpr\u00e1vy CrowdStrike 2025 Global Threat Report do\u0161lo mezi prvn\u00ed a druhou polovinou roku 2024 k n\u00e1r\u016fstu hlasov\u00fdch phishingov\u00fdch (vishingov\u00fdch) \u00fatok\u016f o 442%, a to d\u00edky taktice phishingu a vyd\u00e1v\u00e1n\u00ed se za podvodn\u00edky generovan\u00e9 um\u011blou inteligenc\u00ed.<\/p>\n\n\n\n
Bud\u00ed\u010dek pro $25 milion\u016f<\/strong><\/h2>\n\n\n\n<\/figure>\n<\/div>\n\n\nD\u016fsledky \u00fatok\u016f s vyu\u017eit\u00edm deepfake sahaj\u00ed daleko za hranice spole\u010dnosti Salesforce. <\/p>\n\n\n\n
Lupi\u010dsk\u00e1 loupe\u017e ve v\u00fd\u0161i $25 milion\u016f ve stroj\u00edrensk\u00e9 firm\u011b Arup na za\u010d\u00e1tku roku 2024, kdy \u00fato\u010dn\u00edci pou\u017eili um\u011blou inteligenci, aby se vyd\u00e1vali za n\u011bkolik vedouc\u00edch pracovn\u00edk\u016f ve videohovoru, uk\u00e1zala, \u017ee \u017e\u00e1dn\u00e1 organizace nen\u00ed v\u016f\u010di t\u00e9to hrozb\u011b imunn\u00ed. <\/p>\n\n\n\n
Podobn\u00e9 \u00fatoky se zam\u011b\u0159ily tak\u00e9 na vedouc\u00ed pracovn\u00edky v r\u016fzn\u00fdch odv\u011btv\u00edch, v\u010detn\u011b pokusu vyd\u00e1vat se za gener\u00e1ln\u00edho \u0159editele spole\u010dnosti Ferrari Benedetta Vignu pomoc\u00ed hlasov\u00fdch hovor\u016f s um\u011blou inteligenc\u00ed, kter\u00e9 napodobovaly jeho jihoitalsk\u00fd p\u0159\u00edzvuk.<\/p>\n\n\n\n
Tyto incidenty p\u0159edstavuj\u00ed to, co bezpe\u010dnostn\u00ed experti naz\u00fdvaj\u00ed \"podvody na \u0159editele 2.0\", tedy \u00fatoky, kter\u00e9 jdou nad r\u00e1mec prost\u00e9ho vyd\u00e1v\u00e1n\u00ed se za e-mail a vytv\u00e1\u0159ej\u00ed v\u00edcesmyslov\u00e9 podvody, kter\u00e9 mohou oklamat i zku\u0161en\u00e9 vedouc\u00ed pracovn\u00edky.<\/p>\n\n\n\n
Bezpe\u010dnost platformy vs. lidsk\u00e1 zranitelnost<\/strong><\/h2>\n\n\n\nSpole\u010dnost Salesforce rychle zd\u016fraznila, \u017ee tato naru\u0161en\u00ed nep\u0159edstavuj\u00ed zranitelnost jej\u00ed platformy. <\/p>\n\n\n\n
Spole\u010dnost Salesforce kampa\u0148 UNC6040 p\u0159iznala v b\u0159eznu 2025 a varovala, \u017ee \u00fato\u010dn\u00edci se vyd\u00e1vaj\u00ed za IT podporu a sna\u017e\u00ed se p\u0159im\u011bt zam\u011bstnance, aby jim p\u0159edali p\u0159ihla\u0161ovac\u00ed \u00fadaje nebo schv\u00e1lili \u0161kodliv\u00e9 p\u0159ipojen\u00e9 aplikace. <\/p>\n\n\n\n
Spole\u010dnost zd\u016fraznila, \u017ee tyto incidenty se net\u00fdkaly \u017e\u00e1dn\u00fdch zranitelnost\u00ed v jej\u00ed platform\u011b ani z nich nevypl\u00fdvaly.<\/p>\n\n\n\n
To ukazuje hlavn\u00ed v\u00fdzvu pro v\u0161echny poskytovatele SaaS: vy\u0159e\u0161it, jak zabr\u00e1nit \u00fatok\u016fm, kter\u00e9 hraj\u00ed na lidskou d\u016fv\u011bru m\u00edsto technick\u00fdch chyb.<\/p>\n\n\n\n
Spole\u010dnost Salesforce zavedla n\u011bkolik obrann\u00fdch opat\u0159en\u00ed:<\/p>\n\n\n\n
\n- Zpevn\u011bn\u00ed p\u0159ipojen\u00fdch aplikac\u00ed: Automatick\u00e9 zak\u00e1z\u00e1n\u00ed nenainstalovan\u00fdch p\u0159ipojen\u00fdch aplikac\u00ed pro nov\u00e9 u\u017eivatele<\/li>\n\n\n\n
- Omezen\u00ed toku OAuth: Zak\u00e1z\u00e1n\u00ed p\u0159ipojen\u00ed z\u00edskan\u00fdch pomoc\u00ed ur\u010dit\u00fdch autoriza\u010dn\u00edch proces\u016f<\/li>\n\n\n\n
- Roz\u0161\u00ed\u0159en\u00e9 monitorov\u00e1n\u00ed: Vylep\u0161en\u00e1 detekce podez\u0159el\u00fdch vzor\u016f autorizace aplikac\u00ed<\/li>\n\n\n\n
- Vzd\u011bl\u00e1v\u00e1n\u00ed u\u017eivatel\u016f: Zve\u0159ejn\u011bn\u00ed pokyn\u016f k rozpozn\u00e1v\u00e1n\u00ed pokus\u016f o soci\u00e1ln\u00ed in\u017een\u00fdrstv\u00ed<\/li>\n<\/ul>\n\n\n\n
V srpnu 2025 spole\u010dnost Salesforce uzav\u0159ela v\u0161echny integrace s technologiemi Salesloft, v\u010detn\u011b aplikace Drift, pot\u00e9, co zjistila, \u017ee p\u0159i souvisej\u00edc\u00edch \u00fatoc\u00edch byly odcizeny tokeny OAuth. <\/p>\n\n\n\n
Bezpe\u010dnostn\u00ed experti nyn\u00ed varuj\u00ed, \u017ee t\u00fdmy mus\u00ed naslouchat jemn\u00fdm n\u00e1znak\u016fm, jako jsou zvl\u00e1\u0161tn\u00ed pauzy, \u0161um v pozad\u00ed nebo zvukov\u00e9 z\u00e1vady, kter\u00e9 mohou odhalit, \u017ee hlas byl vytvo\u0159en um\u011blou inteligenc\u00ed.<\/p>\n\n\n\n
Reakce podniku: Nad r\u00e1mec technologick\u00fdch \u0159e\u0161en\u00ed<\/strong><\/h2>\n\n\n\nOrganizace si za\u010d\u00ednaj\u00ed uv\u011bdomovat, \u017ee obrana proti \u00fatok\u016fm s vyu\u017eit\u00edm deepfake vy\u017eaduje v\u00edce ne\u017e jen technologick\u00e1 \u0159e\u0161en\u00ed; vy\u017eaduje z\u00e1sadn\u00ed p\u0159ehodnocen\u00ed proces\u016f d\u016fv\u011bryhodnosti a ov\u011b\u0159ov\u00e1n\u00ed.<\/p>\n\n\n\n
Komunikace s nulovou d\u016fv\u011brou<\/strong><\/h3>\n\n\n\nRealityCheck od spole\u010dnosti Beyond Identity poskytuje ka\u017ed\u00e9mu \u00fa\u010dastn\u00edkovi viditeln\u00fd, ov\u011b\u0159en\u00fd identifika\u010dn\u00ed odznak, kter\u00fd je podpo\u0159en kryptografick\u00fdm ov\u011b\u0159ov\u00e1n\u00edm za\u0159\u00edzen\u00ed a pr\u016fb\u011b\u017enou kontrolou rizik, v sou\u010dasn\u00e9 dob\u011b dostupn\u00fd pro Zoom a Microsoft Teams. <\/p>\n\n\n\n
Takov\u00e1 \u0159e\u0161en\u00ed p\u0159edstavuj\u00ed posun sm\u011brem k model\u016fm \"nikdy ned\u016fv\u011b\u0159uj, v\u017edy ov\u011b\u0159uj\" pro podnikovou komunikaci.<\/p>\n\n\n\n
Roz\u0161\u00ed\u0159en\u00e9 \u0161kolic\u00ed programy<\/strong><\/h3>\n\n\n\nOrganizace, kter\u00e9 pou\u017e\u00edvaj\u00ed platformu pro simulaci deepfake od spole\u010dnosti Resemble AI, hl\u00e1s\u00ed a\u017e o 90% ni\u017e\u0161\u00ed po\u010det \u00fasp\u011b\u0161n\u00fdch \u00fatok\u016f po zaveden\u00ed platformy, kter\u00e1 vyu\u017e\u00edv\u00e1 hyperrealistick\u00e9 simulace k ilustraci toho, jak se deepfake \u00fatoky odehr\u00e1vaj\u00ed v re\u00e1ln\u00e9m sv\u011bt\u011b.<\/p>\n\n\n\n
V\u00edcekan\u00e1lov\u00e9 ov\u011b\u0159ov\u00e1n\u00ed<\/strong><\/h3>\n\n\n\nP\u0159edn\u00ed organizace zav\u00e1d\u011bj\u00ed protokoly, kter\u00e9 vy\u017eaduj\u00ed ov\u011b\u0159en\u00ed v\u00edce kan\u00e1ly pro v\u0161echny vysoce rizikov\u00e9 po\u017eadavky, bez ohledu na to, jak autentick\u00e1 se zd\u00e1 b\u00fdt p\u016fvodn\u00ed komunikace.<\/p>\n\n\n\n
Kdy\u017e u\u017e hlas nen\u00ed pravda<\/strong><\/h2>\n\n\n\nHrozba deepfake pro Salesforce a dal\u0161\u00ed podnikov\u00e9 syst\u00e9my p\u0159edstavuje zlomov\u00fd bod v historii kybernetick\u00e9 bezpe\u010dnosti. <\/p>\n\n\n\n
Poprv\u00e9 se \u00fato\u010dn\u00edci mohou nejen vyd\u00e1vat za autority, ale mohou za n\u011b i mluvit, vypadat jako oni a p\u0159esv\u011bd\u010dit i vy\u0161kolen\u00e9 bezpe\u010dnostn\u00ed profesion\u00e1ly, aby podnikli kroky, kter\u00e9 ohro\u017euj\u00ed bezpe\u010dnost organizace.<\/p>\n\n\n\n
V dob\u011b podvod\u016f generovan\u00fdch um\u011blou inteligenc\u00ed je t\u0159eba si d\u016fv\u011bru zaslou\u017eit ov\u011b\u0159en\u00edm, nikoliv p\u0159edpokl\u00e1dat, \u017ee je d\u016fv\u011bryhodn\u00e1.<\/p>\n\n\n\n
Spole\u010dnosti, kter\u00e9 si to uv\u011bdomuj\u00ed a investuj\u00ed do spr\u00e1vn\u00fdch n\u00e1stroj\u016f, proces\u016f a kultury, budou m\u00edt nejlep\u0161\u00ed pozici pro ochranu bezpe\u010dnosti i d\u016fv\u011bry ve st\u00e1le nejist\u011bj\u0161\u00edm digit\u00e1ln\u00edm sv\u011bt\u011b.<\/p>\n\n\n\n
V t\u00e9to nov\u00e9 realit\u011b nemus\u00ed b\u00fdt hlas na druh\u00e9m konci linky t\u00edm, k\u00fdm se zd\u00e1.<\/p>\n\n\n\n
V dob\u011b deepfakes je neust\u00e1l\u00e1 ostra\u017eitost cenou za bezpe\u010dnost.<\/p>","protected":false},"excerpt":{"rendered":"
V \u010dervnu 2025 obdr\u017eel zam\u011bstnanec spole\u010dnosti Google zd\u00e1nliv\u011b rutinn\u00ed telefon\u00e1t od [...]<\/p>","protected":false},"author":15,"featured_media":5339,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"rank_math_lock_modified_date":false,"_themeisle_gutenberg_block_has_review":false,"footnotes":""},"categories":[31],"tags":[],"class_list":["post-5335","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-helpful-ai-content-tips"],"_links":{"self":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/users\/15"}],"replies":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/comments?post=5335"}],"version-history":[{"count":3,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335\/revisions"}],"predecessor-version":[{"id":5343,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335\/revisions\/5343"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/media\/5339"}],"wp:attachment":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/media?parent=5335"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/categories?post=5335"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/tags?post=5335"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
<\/figure>\n<\/div>\n\n\nD\u016fsledky \u00fatok\u016f s vyu\u017eit\u00edm deepfake sahaj\u00ed daleko za hranice spole\u010dnosti Salesforce. <\/p>\n\n\n\n
Lupi\u010dsk\u00e1 loupe\u017e ve v\u00fd\u0161i $25 milion\u016f ve stroj\u00edrensk\u00e9 firm\u011b Arup na za\u010d\u00e1tku roku 2024, kdy \u00fato\u010dn\u00edci pou\u017eili um\u011blou inteligenci, aby se vyd\u00e1vali za n\u011bkolik vedouc\u00edch pracovn\u00edk\u016f ve videohovoru, uk\u00e1zala, \u017ee \u017e\u00e1dn\u00e1 organizace nen\u00ed v\u016f\u010di t\u00e9to hrozb\u011b imunn\u00ed. <\/p>\n\n\n\n
Podobn\u00e9 \u00fatoky se zam\u011b\u0159ily tak\u00e9 na vedouc\u00ed pracovn\u00edky v r\u016fzn\u00fdch odv\u011btv\u00edch, v\u010detn\u011b pokusu vyd\u00e1vat se za gener\u00e1ln\u00edho \u0159editele spole\u010dnosti Ferrari Benedetta Vignu pomoc\u00ed hlasov\u00fdch hovor\u016f s um\u011blou inteligenc\u00ed, kter\u00e9 napodobovaly jeho jihoitalsk\u00fd p\u0159\u00edzvuk.<\/p>\n\n\n\n
Tyto incidenty p\u0159edstavuj\u00ed to, co bezpe\u010dnostn\u00ed experti naz\u00fdvaj\u00ed \"podvody na \u0159editele 2.0\", tedy \u00fatoky, kter\u00e9 jdou nad r\u00e1mec prost\u00e9ho vyd\u00e1v\u00e1n\u00ed se za e-mail a vytv\u00e1\u0159ej\u00ed v\u00edcesmyslov\u00e9 podvody, kter\u00e9 mohou oklamat i zku\u0161en\u00e9 vedouc\u00ed pracovn\u00edky.<\/p>\n\n\n\n
Bezpe\u010dnost platformy vs. lidsk\u00e1 zranitelnost<\/strong><\/h2>\n\n\n\nSpole\u010dnost Salesforce rychle zd\u016fraznila, \u017ee tato naru\u0161en\u00ed nep\u0159edstavuj\u00ed zranitelnost jej\u00ed platformy. <\/p>\n\n\n\n
Spole\u010dnost Salesforce kampa\u0148 UNC6040 p\u0159iznala v b\u0159eznu 2025 a varovala, \u017ee \u00fato\u010dn\u00edci se vyd\u00e1vaj\u00ed za IT podporu a sna\u017e\u00ed se p\u0159im\u011bt zam\u011bstnance, aby jim p\u0159edali p\u0159ihla\u0161ovac\u00ed \u00fadaje nebo schv\u00e1lili \u0161kodliv\u00e9 p\u0159ipojen\u00e9 aplikace. <\/p>\n\n\n\n
Spole\u010dnost zd\u016fraznila, \u017ee tyto incidenty se net\u00fdkaly \u017e\u00e1dn\u00fdch zranitelnost\u00ed v jej\u00ed platform\u011b ani z nich nevypl\u00fdvaly.<\/p>\n\n\n\n
To ukazuje hlavn\u00ed v\u00fdzvu pro v\u0161echny poskytovatele SaaS: vy\u0159e\u0161it, jak zabr\u00e1nit \u00fatok\u016fm, kter\u00e9 hraj\u00ed na lidskou d\u016fv\u011bru m\u00edsto technick\u00fdch chyb.<\/p>\n\n\n\n
Spole\u010dnost Salesforce zavedla n\u011bkolik obrann\u00fdch opat\u0159en\u00ed:<\/p>\n\n\n\n
\n- Zpevn\u011bn\u00ed p\u0159ipojen\u00fdch aplikac\u00ed: Automatick\u00e9 zak\u00e1z\u00e1n\u00ed nenainstalovan\u00fdch p\u0159ipojen\u00fdch aplikac\u00ed pro nov\u00e9 u\u017eivatele<\/li>\n\n\n\n
- Omezen\u00ed toku OAuth: Zak\u00e1z\u00e1n\u00ed p\u0159ipojen\u00ed z\u00edskan\u00fdch pomoc\u00ed ur\u010dit\u00fdch autoriza\u010dn\u00edch proces\u016f<\/li>\n\n\n\n
- Roz\u0161\u00ed\u0159en\u00e9 monitorov\u00e1n\u00ed: Vylep\u0161en\u00e1 detekce podez\u0159el\u00fdch vzor\u016f autorizace aplikac\u00ed<\/li>\n\n\n\n
- Vzd\u011bl\u00e1v\u00e1n\u00ed u\u017eivatel\u016f: Zve\u0159ejn\u011bn\u00ed pokyn\u016f k rozpozn\u00e1v\u00e1n\u00ed pokus\u016f o soci\u00e1ln\u00ed in\u017een\u00fdrstv\u00ed<\/li>\n<\/ul>\n\n\n\n
V srpnu 2025 spole\u010dnost Salesforce uzav\u0159ela v\u0161echny integrace s technologiemi Salesloft, v\u010detn\u011b aplikace Drift, pot\u00e9, co zjistila, \u017ee p\u0159i souvisej\u00edc\u00edch \u00fatoc\u00edch byly odcizeny tokeny OAuth. <\/p>\n\n\n\n
Bezpe\u010dnostn\u00ed experti nyn\u00ed varuj\u00ed, \u017ee t\u00fdmy mus\u00ed naslouchat jemn\u00fdm n\u00e1znak\u016fm, jako jsou zvl\u00e1\u0161tn\u00ed pauzy, \u0161um v pozad\u00ed nebo zvukov\u00e9 z\u00e1vady, kter\u00e9 mohou odhalit, \u017ee hlas byl vytvo\u0159en um\u011blou inteligenc\u00ed.<\/p>\n\n\n\n
Reakce podniku: Nad r\u00e1mec technologick\u00fdch \u0159e\u0161en\u00ed<\/strong><\/h2>\n\n\n\nOrganizace si za\u010d\u00ednaj\u00ed uv\u011bdomovat, \u017ee obrana proti \u00fatok\u016fm s vyu\u017eit\u00edm deepfake vy\u017eaduje v\u00edce ne\u017e jen technologick\u00e1 \u0159e\u0161en\u00ed; vy\u017eaduje z\u00e1sadn\u00ed p\u0159ehodnocen\u00ed proces\u016f d\u016fv\u011bryhodnosti a ov\u011b\u0159ov\u00e1n\u00ed.<\/p>\n\n\n\n
Komunikace s nulovou d\u016fv\u011brou<\/strong><\/h3>\n\n\n\nRealityCheck od spole\u010dnosti Beyond Identity poskytuje ka\u017ed\u00e9mu \u00fa\u010dastn\u00edkovi viditeln\u00fd, ov\u011b\u0159en\u00fd identifika\u010dn\u00ed odznak, kter\u00fd je podpo\u0159en kryptografick\u00fdm ov\u011b\u0159ov\u00e1n\u00edm za\u0159\u00edzen\u00ed a pr\u016fb\u011b\u017enou kontrolou rizik, v sou\u010dasn\u00e9 dob\u011b dostupn\u00fd pro Zoom a Microsoft Teams. <\/p>\n\n\n\n
Takov\u00e1 \u0159e\u0161en\u00ed p\u0159edstavuj\u00ed posun sm\u011brem k model\u016fm \"nikdy ned\u016fv\u011b\u0159uj, v\u017edy ov\u011b\u0159uj\" pro podnikovou komunikaci.<\/p>\n\n\n\n
Roz\u0161\u00ed\u0159en\u00e9 \u0161kolic\u00ed programy<\/strong><\/h3>\n\n\n\nOrganizace, kter\u00e9 pou\u017e\u00edvaj\u00ed platformu pro simulaci deepfake od spole\u010dnosti Resemble AI, hl\u00e1s\u00ed a\u017e o 90% ni\u017e\u0161\u00ed po\u010det \u00fasp\u011b\u0161n\u00fdch \u00fatok\u016f po zaveden\u00ed platformy, kter\u00e1 vyu\u017e\u00edv\u00e1 hyperrealistick\u00e9 simulace k ilustraci toho, jak se deepfake \u00fatoky odehr\u00e1vaj\u00ed v re\u00e1ln\u00e9m sv\u011bt\u011b.<\/p>\n\n\n\n
V\u00edcekan\u00e1lov\u00e9 ov\u011b\u0159ov\u00e1n\u00ed<\/strong><\/h3>\n\n\n\nP\u0159edn\u00ed organizace zav\u00e1d\u011bj\u00ed protokoly, kter\u00e9 vy\u017eaduj\u00ed ov\u011b\u0159en\u00ed v\u00edce kan\u00e1ly pro v\u0161echny vysoce rizikov\u00e9 po\u017eadavky, bez ohledu na to, jak autentick\u00e1 se zd\u00e1 b\u00fdt p\u016fvodn\u00ed komunikace.<\/p>\n\n\n\n
Kdy\u017e u\u017e hlas nen\u00ed pravda<\/strong><\/h2>\n\n\n\nHrozba deepfake pro Salesforce a dal\u0161\u00ed podnikov\u00e9 syst\u00e9my p\u0159edstavuje zlomov\u00fd bod v historii kybernetick\u00e9 bezpe\u010dnosti. <\/p>\n\n\n\n
Poprv\u00e9 se \u00fato\u010dn\u00edci mohou nejen vyd\u00e1vat za autority, ale mohou za n\u011b i mluvit, vypadat jako oni a p\u0159esv\u011bd\u010dit i vy\u0161kolen\u00e9 bezpe\u010dnostn\u00ed profesion\u00e1ly, aby podnikli kroky, kter\u00e9 ohro\u017euj\u00ed bezpe\u010dnost organizace.<\/p>\n\n\n\n
V dob\u011b podvod\u016f generovan\u00fdch um\u011blou inteligenc\u00ed je t\u0159eba si d\u016fv\u011bru zaslou\u017eit ov\u011b\u0159en\u00edm, nikoliv p\u0159edpokl\u00e1dat, \u017ee je d\u016fv\u011bryhodn\u00e1.<\/p>\n\n\n\n
Spole\u010dnosti, kter\u00e9 si to uv\u011bdomuj\u00ed a investuj\u00ed do spr\u00e1vn\u00fdch n\u00e1stroj\u016f, proces\u016f a kultury, budou m\u00edt nejlep\u0161\u00ed pozici pro ochranu bezpe\u010dnosti i d\u016fv\u011bry ve st\u00e1le nejist\u011bj\u0161\u00edm digit\u00e1ln\u00edm sv\u011bt\u011b.<\/p>\n\n\n\n
V t\u00e9to nov\u00e9 realit\u011b nemus\u00ed b\u00fdt hlas na druh\u00e9m konci linky t\u00edm, k\u00fdm se zd\u00e1.<\/p>\n\n\n\n
V dob\u011b deepfakes je neust\u00e1l\u00e1 ostra\u017eitost cenou za bezpe\u010dnost.<\/p>","protected":false},"excerpt":{"rendered":"
V \u010dervnu 2025 obdr\u017eel zam\u011bstnanec spole\u010dnosti Google zd\u00e1nliv\u011b rutinn\u00ed telefon\u00e1t od [...]<\/p>","protected":false},"author":15,"featured_media":5339,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"rank_math_lock_modified_date":false,"_themeisle_gutenberg_block_has_review":false,"footnotes":""},"categories":[31],"tags":[],"class_list":["post-5335","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-helpful-ai-content-tips"],"_links":{"self":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/users\/15"}],"replies":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/comments?post=5335"}],"version-history":[{"count":3,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335\/revisions"}],"predecessor-version":[{"id":5343,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335\/revisions\/5343"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/media\/5339"}],"wp:attachment":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/media?parent=5335"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/categories?post=5335"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/tags?post=5335"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
V srpnu 2025 spole\u010dnost Salesforce uzav\u0159ela v\u0161echny integrace s technologiemi Salesloft, v\u010detn\u011b aplikace Drift, pot\u00e9, co zjistila, \u017ee p\u0159i souvisej\u00edc\u00edch \u00fatoc\u00edch byly odcizeny tokeny OAuth. <\/p>\n\n\n\n
Bezpe\u010dnostn\u00ed experti nyn\u00ed varuj\u00ed, \u017ee t\u00fdmy mus\u00ed naslouchat jemn\u00fdm n\u00e1znak\u016fm, jako jsou zvl\u00e1\u0161tn\u00ed pauzy, \u0161um v pozad\u00ed nebo zvukov\u00e9 z\u00e1vady, kter\u00e9 mohou odhalit, \u017ee hlas byl vytvo\u0159en um\u011blou inteligenc\u00ed.<\/p>\n\n\n\n
Reakce podniku: Nad r\u00e1mec technologick\u00fdch \u0159e\u0161en\u00ed<\/strong><\/h2>\n\n\n\nOrganizace si za\u010d\u00ednaj\u00ed uv\u011bdomovat, \u017ee obrana proti \u00fatok\u016fm s vyu\u017eit\u00edm deepfake vy\u017eaduje v\u00edce ne\u017e jen technologick\u00e1 \u0159e\u0161en\u00ed; vy\u017eaduje z\u00e1sadn\u00ed p\u0159ehodnocen\u00ed proces\u016f d\u016fv\u011bryhodnosti a ov\u011b\u0159ov\u00e1n\u00ed.<\/p>\n\n\n\n
Komunikace s nulovou d\u016fv\u011brou<\/strong><\/h3>\n\n\n\nRealityCheck od spole\u010dnosti Beyond Identity poskytuje ka\u017ed\u00e9mu \u00fa\u010dastn\u00edkovi viditeln\u00fd, ov\u011b\u0159en\u00fd identifika\u010dn\u00ed odznak, kter\u00fd je podpo\u0159en kryptografick\u00fdm ov\u011b\u0159ov\u00e1n\u00edm za\u0159\u00edzen\u00ed a pr\u016fb\u011b\u017enou kontrolou rizik, v sou\u010dasn\u00e9 dob\u011b dostupn\u00fd pro Zoom a Microsoft Teams. <\/p>\n\n\n\n
Takov\u00e1 \u0159e\u0161en\u00ed p\u0159edstavuj\u00ed posun sm\u011brem k model\u016fm \"nikdy ned\u016fv\u011b\u0159uj, v\u017edy ov\u011b\u0159uj\" pro podnikovou komunikaci.<\/p>\n\n\n\n
Roz\u0161\u00ed\u0159en\u00e9 \u0161kolic\u00ed programy<\/strong><\/h3>\n\n\n\nOrganizace, kter\u00e9 pou\u017e\u00edvaj\u00ed platformu pro simulaci deepfake od spole\u010dnosti Resemble AI, hl\u00e1s\u00ed a\u017e o 90% ni\u017e\u0161\u00ed po\u010det \u00fasp\u011b\u0161n\u00fdch \u00fatok\u016f po zaveden\u00ed platformy, kter\u00e1 vyu\u017e\u00edv\u00e1 hyperrealistick\u00e9 simulace k ilustraci toho, jak se deepfake \u00fatoky odehr\u00e1vaj\u00ed v re\u00e1ln\u00e9m sv\u011bt\u011b.<\/p>\n\n\n\n
V\u00edcekan\u00e1lov\u00e9 ov\u011b\u0159ov\u00e1n\u00ed<\/strong><\/h3>\n\n\n\nP\u0159edn\u00ed organizace zav\u00e1d\u011bj\u00ed protokoly, kter\u00e9 vy\u017eaduj\u00ed ov\u011b\u0159en\u00ed v\u00edce kan\u00e1ly pro v\u0161echny vysoce rizikov\u00e9 po\u017eadavky, bez ohledu na to, jak autentick\u00e1 se zd\u00e1 b\u00fdt p\u016fvodn\u00ed komunikace.<\/p>\n\n\n\n
Kdy\u017e u\u017e hlas nen\u00ed pravda<\/strong><\/h2>\n\n\n\nHrozba deepfake pro Salesforce a dal\u0161\u00ed podnikov\u00e9 syst\u00e9my p\u0159edstavuje zlomov\u00fd bod v historii kybernetick\u00e9 bezpe\u010dnosti. <\/p>\n\n\n\n
Poprv\u00e9 se \u00fato\u010dn\u00edci mohou nejen vyd\u00e1vat za autority, ale mohou za n\u011b i mluvit, vypadat jako oni a p\u0159esv\u011bd\u010dit i vy\u0161kolen\u00e9 bezpe\u010dnostn\u00ed profesion\u00e1ly, aby podnikli kroky, kter\u00e9 ohro\u017euj\u00ed bezpe\u010dnost organizace.<\/p>\n\n\n\n
V dob\u011b podvod\u016f generovan\u00fdch um\u011blou inteligenc\u00ed je t\u0159eba si d\u016fv\u011bru zaslou\u017eit ov\u011b\u0159en\u00edm, nikoliv p\u0159edpokl\u00e1dat, \u017ee je d\u016fv\u011bryhodn\u00e1.<\/p>\n\n\n\n
Spole\u010dnosti, kter\u00e9 si to uv\u011bdomuj\u00ed a investuj\u00ed do spr\u00e1vn\u00fdch n\u00e1stroj\u016f, proces\u016f a kultury, budou m\u00edt nejlep\u0161\u00ed pozici pro ochranu bezpe\u010dnosti i d\u016fv\u011bry ve st\u00e1le nejist\u011bj\u0161\u00edm digit\u00e1ln\u00edm sv\u011bt\u011b.<\/p>\n\n\n\n
V t\u00e9to nov\u00e9 realit\u011b nemus\u00ed b\u00fdt hlas na druh\u00e9m konci linky t\u00edm, k\u00fdm se zd\u00e1.<\/p>\n\n\n\n
V dob\u011b deepfakes je neust\u00e1l\u00e1 ostra\u017eitost cenou za bezpe\u010dnost.<\/p>","protected":false},"excerpt":{"rendered":"
V \u010dervnu 2025 obdr\u017eel zam\u011bstnanec spole\u010dnosti Google zd\u00e1nliv\u011b rutinn\u00ed telefon\u00e1t od [...]<\/p>","protected":false},"author":15,"featured_media":5339,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"rank_math_lock_modified_date":false,"_themeisle_gutenberg_block_has_review":false,"footnotes":""},"categories":[31],"tags":[],"class_list":["post-5335","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-helpful-ai-content-tips"],"_links":{"self":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/users\/15"}],"replies":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/comments?post=5335"}],"version-history":[{"count":3,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335\/revisions"}],"predecessor-version":[{"id":5343,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335\/revisions\/5343"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/media\/5339"}],"wp:attachment":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/media?parent=5335"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/categories?post=5335"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/tags?post=5335"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
RealityCheck od spole\u010dnosti Beyond Identity poskytuje ka\u017ed\u00e9mu \u00fa\u010dastn\u00edkovi viditeln\u00fd, ov\u011b\u0159en\u00fd identifika\u010dn\u00ed odznak, kter\u00fd je podpo\u0159en kryptografick\u00fdm ov\u011b\u0159ov\u00e1n\u00edm za\u0159\u00edzen\u00ed a pr\u016fb\u011b\u017enou kontrolou rizik, v sou\u010dasn\u00e9 dob\u011b dostupn\u00fd pro Zoom a Microsoft Teams. <\/p>\n\n\n\n
Takov\u00e1 \u0159e\u0161en\u00ed p\u0159edstavuj\u00ed posun sm\u011brem k model\u016fm \"nikdy ned\u016fv\u011b\u0159uj, v\u017edy ov\u011b\u0159uj\" pro podnikovou komunikaci.<\/p>\n\n\n\n
Roz\u0161\u00ed\u0159en\u00e9 \u0161kolic\u00ed programy<\/strong><\/h3>\n\n\n\nOrganizace, kter\u00e9 pou\u017e\u00edvaj\u00ed platformu pro simulaci deepfake od spole\u010dnosti Resemble AI, hl\u00e1s\u00ed a\u017e o 90% ni\u017e\u0161\u00ed po\u010det \u00fasp\u011b\u0161n\u00fdch \u00fatok\u016f po zaveden\u00ed platformy, kter\u00e1 vyu\u017e\u00edv\u00e1 hyperrealistick\u00e9 simulace k ilustraci toho, jak se deepfake \u00fatoky odehr\u00e1vaj\u00ed v re\u00e1ln\u00e9m sv\u011bt\u011b.<\/p>\n\n\n\n
V\u00edcekan\u00e1lov\u00e9 ov\u011b\u0159ov\u00e1n\u00ed<\/strong><\/h3>\n\n\n\nP\u0159edn\u00ed organizace zav\u00e1d\u011bj\u00ed protokoly, kter\u00e9 vy\u017eaduj\u00ed ov\u011b\u0159en\u00ed v\u00edce kan\u00e1ly pro v\u0161echny vysoce rizikov\u00e9 po\u017eadavky, bez ohledu na to, jak autentick\u00e1 se zd\u00e1 b\u00fdt p\u016fvodn\u00ed komunikace.<\/p>\n\n\n\n
Kdy\u017e u\u017e hlas nen\u00ed pravda<\/strong><\/h2>\n\n\n\nHrozba deepfake pro Salesforce a dal\u0161\u00ed podnikov\u00e9 syst\u00e9my p\u0159edstavuje zlomov\u00fd bod v historii kybernetick\u00e9 bezpe\u010dnosti. <\/p>\n\n\n\n
Poprv\u00e9 se \u00fato\u010dn\u00edci mohou nejen vyd\u00e1vat za autority, ale mohou za n\u011b i mluvit, vypadat jako oni a p\u0159esv\u011bd\u010dit i vy\u0161kolen\u00e9 bezpe\u010dnostn\u00ed profesion\u00e1ly, aby podnikli kroky, kter\u00e9 ohro\u017euj\u00ed bezpe\u010dnost organizace.<\/p>\n\n\n\n
V dob\u011b podvod\u016f generovan\u00fdch um\u011blou inteligenc\u00ed je t\u0159eba si d\u016fv\u011bru zaslou\u017eit ov\u011b\u0159en\u00edm, nikoliv p\u0159edpokl\u00e1dat, \u017ee je d\u016fv\u011bryhodn\u00e1.<\/p>\n\n\n\n
Spole\u010dnosti, kter\u00e9 si to uv\u011bdomuj\u00ed a investuj\u00ed do spr\u00e1vn\u00fdch n\u00e1stroj\u016f, proces\u016f a kultury, budou m\u00edt nejlep\u0161\u00ed pozici pro ochranu bezpe\u010dnosti i d\u016fv\u011bry ve st\u00e1le nejist\u011bj\u0161\u00edm digit\u00e1ln\u00edm sv\u011bt\u011b.<\/p>\n\n\n\n
V t\u00e9to nov\u00e9 realit\u011b nemus\u00ed b\u00fdt hlas na druh\u00e9m konci linky t\u00edm, k\u00fdm se zd\u00e1.<\/p>\n\n\n\n
V dob\u011b deepfakes je neust\u00e1l\u00e1 ostra\u017eitost cenou za bezpe\u010dnost.<\/p>","protected":false},"excerpt":{"rendered":"
V \u010dervnu 2025 obdr\u017eel zam\u011bstnanec spole\u010dnosti Google zd\u00e1nliv\u011b rutinn\u00ed telefon\u00e1t od [...]<\/p>","protected":false},"author":15,"featured_media":5339,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"rank_math_lock_modified_date":false,"_themeisle_gutenberg_block_has_review":false,"footnotes":""},"categories":[31],"tags":[],"class_list":["post-5335","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-helpful-ai-content-tips"],"_links":{"self":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/users\/15"}],"replies":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/comments?post=5335"}],"version-history":[{"count":3,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335\/revisions"}],"predecessor-version":[{"id":5343,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/posts\/5335\/revisions\/5343"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/media\/5339"}],"wp:attachment":[{"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/media?parent=5335"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/categories?post=5335"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blog.truthscan.com\/cs\/wp-json\/wp\/v2\/tags?post=5335"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
P\u0159edn\u00ed organizace zav\u00e1d\u011bj\u00ed protokoly, kter\u00e9 vy\u017eaduj\u00ed ov\u011b\u0159en\u00ed v\u00edce kan\u00e1ly pro v\u0161echny vysoce rizikov\u00e9 po\u017eadavky, bez ohledu na to, jak autentick\u00e1 se zd\u00e1 b\u00fdt p\u016fvodn\u00ed komunikace.<\/p>\n\n\n\n