Kwenzekeni
Umbiko we-Cloud Security Alliance’s 2026 Top Threats to Cloud Computing uhlonza izinkinga ezingu-11 ezibalulekile zokuphepha kwamafu ngokusekelwe kuhlolo lomhlaba wonke lochwepheshe bemboni. Ilinganisa ubunikazi obungenele nokuphathwa kokufinyelela njengosongo oluhamba phambili futhi yengeza ukuhlaselwa okuthuthukisiwe kwe-AI kanye nokuyekethisa kohlelo lwe-AI kumazinga okokuqala ngqa.
I-Cloud Security Alliance yethula i-2026 Top Threats to Cloud Computing Survey Report njengokuhlolwa kwezindaba eziyi-11 ezibaluleke kakhulu zokuphepha kwamafu ezikhonjwe ngocwaningo lomhlaba wonke lochwepheshe bemboni. Ikhasi lomthombo lichaza umbiko njengobonisa ushintsho olubalulekile ezintweni ezibalulekile zokuphepha kwamafu phakathi neminyaka emibili edlule. Ikhasi alinikezi usayizi wesampula yenhlolovo, amadethi enkundla, ukuhlukaniswa kwabaphenduli, indlela yokushaya amaphuzu noma umkhawulo wephutha, ngakho leyo mininingwane ihlala ingaziwa kulo mthombo.
Umbiko ubeka ubunikazi obungenele nokuphathwa kokufinyelela phezulu ohlwini lwawo lwango-2026. Ngokusho kwe-CSA, lokho kuqala kubonisa izingozi ezibandakanya izimvume ezeqile, ubunikazi obungebona abantu, imininingwane ephethwe kabi kanye nobudlelwano bokuthembana obuhlanganisiwe. Umthombo awunikezi amaphuzu ezinombolo, ukuqhathaniswa kwezinga nama-edishini angaphambili noma izehlakalo ezithile ezisekela ukubekwa. Lokho okweqiwe kwenza isiqondiso sokutholakele sicace kuyilapho kushiya ubukhulu bakho kanye nokugcwaliseka kwakho kungaqinisekile.
Izingozi ezimbili ezihlobene ne-AI zingena kumazinga okokuqala ngqa. Umbiko uchaza ukuhlasela okuthuthukisiwe kwe-AI njengezindlela izitha ezisebenzisa i-AI ukuze zithuthukise futhi zizenzele ukuhlasela. Ichaza ngokwehlukana ukonakala kohlelo lwe-AI njengokukhohlisa noma ukuhlukumeza amamodeli we-AI, idatha, ama-ejenti, amathuluzi namapayipi. Lo mehluko wenza i-AI ibe yisihloko esiqondile sokutholwa kwezokuphepha: isigaba esisodwa siphathelene ne-AI esetshenziswa abahlaseli, kanti esinye siphathelene nokuhlaselwa kwamasistimu akhiwe eduze kwe-AI.
I-CSA ithi umbiko uhlola imithelela yezobuchwepheshe nebhizinisi, izibonelo zomhlaba wangempela kanye nokunciphisa okungokoqobo kosongo ngalunye. Iphinde ithi ukuhlaziya kubeka izinkombandlela ezifanele nezilawuli ezivela ku-CSA Security Guidance v5 kanye ne-AI Cloud Controls Matrix v1.1. Ikhasi elibonakalayo lomthombo alizibandakanyi lezo zibonelo, izilawuli noma izifundo eziyisibonelo, futhi aliqinisekisi ukuthi okutholwe yinhlolovo yombiko kumelela ukuvama kwezigameko eziqinisekisiwe ngokuzimele.
Imininingwane yomthombo: cloudsecurityalliance.org ↗
Kungani kubalulekile
Umbiko ufaka ukuphepha kwe-AI njengengcuphe yamafu ezinhlangothi ezimbili: izitha zingasebenzisa i-AI ukuze zithuthukise noma zizenzele ukuhlasela, kuyilapho amamodeli e-AI, idatha, ama-ejenti, amathuluzi namapayipi ngokwawo angashintshwa noma ahlukunyezwe.
Ukubaluleka okuphakathi kwalo mbiko ukuphatha kwayo i-AI njengamandla angaqinisa ukuhlasela kanye nesigaba esikhulayo sezinhlelo ezidinga ukuvikelwa. Lolo hlaka lucaciswe kakhulu kunokuphatha i-AI njengenkinga yobuchwepheshe ejwayelekile. Ihlukanisa izingozi ezivela ekusetshenzisweni komhlaseli kwe-AI kusukela ezingozini ezidalwe ubuthakathaka kumamodeli, idatha, ama-ejenti, amathuluzi namapayipi izinhlangano ezisebenza ezindaweni zamafu.
Ukumataniswa kobunikazi nobungozi be-AI nakho kukhomba ekuxhumekeni kokusebenza. Abenzeli be-AI nobunye ubunikazi obungebona abantu bangase bathole izimvume, basebenzise amathuluzi noma bahlanganyele nezinsiza zamafu, kuyilapho amapayipi e-AI encike kudatha, isofthiwe nezinsizakalo zezinkampani zangaphandle. Umthombo awusho ukuthi lezi zingozi zihlala zenzeka ndawonye, kodwa izigaba ziphakamisa ukuthi ukuphathwa kokufinyelela nokuvikeleka kwe-AI kungase kudingeke kuhlolwe ndawonye esikhundleni sokuphathwa njengezinhlelo ezihluke ngokuphelele.
Emaqenjini ezokuphepha, umbiko unikeza uhlaka lokubuza imibuzo ebambekayo: ukuthi ingabe ubunikazi obungebona abantu banokufinyelela okwengeziwe kunalokho okudingekile, noma ngabe izifakazelo nobudlelwano bokwethembana buyalawulwa, nokuthi amamodeli we-AI, idatha, ama-ejenti, amathuluzi namapayipi angagadwa ukuze asetshenziswe noma ahlukunyezwe. Lena imithelela engokoqobo yezigaba ze-CSA, akubona ubufakazi bokuthi noma iyiphi inhlangano ethile idaluliwe noma ukuthi ukulawula okuthile kuzovimbela ukuhlasela.
Umbiko ungase ube usizo kuma-CISO, abaklami bamafu, ochwepheshe bezokuphepha be-AI, amaqembu okuphatha nabaphenduli bezehlakalo ngoba umthombo ukhomba ngokuqondile lawo maqembu njengezithameli ezihlosiwe. Imikhawulo yayo ibaluleke ngokufanayo. Ikhasi aliqinisekisi ukuvama, ubukhali noma umthelela wezezimali wezigameko zamafu ezihlobene ne-AI, futhi aliziqinisekisi ngokuzimele izahlulelo zenhlolovo ngokumelene nesizindalwazi sezigameko esibanzi.
I-Interactive Mechanism: Indlela Esebenza Ngayo Ngempela
Hlola ubuchwepheshe obuyisisekelo ngemuva kwalokhu kuthuthukiswa ngokuhlanganyela.
crm_get_transaction(id='4092').Why can ethical evaluation not be reduced to one model score?
Ongakubuka ngokulandelayo
Indlela yombiko ephelele, idatha yezinga, izibonelo kanye nesiqondiso sokunciphisa kuzonquma ukuthi iziphetho zawo zingasetshenziswa kangakanani. Izinhlangano kufanele futhi zibheke ukuthi ubungozi be-AI obuhlonziwe buvela ezenzakalweni ezibhaliwe nokuthi ubunikazi obukhona, izilawuli zefu kanye ne-AI zibhekana kanjani nazo.
Iphoyinti lokuqinisekisa elilandelayo elibaluleke kakhulu umbiko ophelele ongadawuniloda. Abafundi kufanele babheke inani lenhlolovo, ukubunjwa kwendawo kanye nobungcweti, amagama emibuzo, ukubalwa kwezimpendulo, ukukala kanye nemibandela esetshenziswa ukukala izinsongo eziyi-11. Ngaphandle kwaleyo mininingwane, umbiko uphathwa kangcono njengokuhlola okusekelwe kuhlolo lwe-CSA kokubalulekile kunesilinganiso esiqinisekile sokuvama kosongo.
Ukuphathwa kombiko kokuhlaselwa okuthuthukisiwe kwe-AI kuqinisekisa ukunakwa okuseduze kwezibonelo zawo kanye nokunciphisa. Umthombo uthi lesi sigaba sihlanganisa izitha ezisebenzisa i-AI ukuthuthukisa nokwenza ukuhlasela ngokuzenzakalela, kodwa asizihlonzi izindlela ezithile zokuhlasela, imikhakha ethintekile, imikhankaso ebhekiwe noma izinguquko ezilinganisiwe emandleni omhlaseli. Lokho okungaziwa kunquma ukuthi isigaba sichaza ubungozi obuvelayo, inkinga yokusebenza esabalele noma kokubili.
Ukuze kuthoniswe i-AI-system, abafundi kufanele bahlole ukuthi i-CSA ihumusha kanjani amagama abanzi njengamamodeli, idatha, ama-ejenti, amathuluzi namapayipi abe yizidingo ezithile zokulawula. Imininingwane ewusizo ingabandakanya izinhlobo zokukhohlisa noma ukuhlukumeza okucatshangelwayo, lapho isibopho sihlala phakathi kwabahlinzeki bamafu namakhasimende, nokuthi izinhlangano kulindeleke ukuthi zithole futhi zisabele kanjani ekuhlulekeni. Ayikho kuleyo mininingwane yokuqaliswa evela ekhasini elibonakalayo.
Lo mbiko ufanele ukulandelelwa izehlakalo ezibhalwe phansi kanye nokuhlolwa kwezinsongo ezizayo. Uma ubufakazi bakamuva bukhombisa ukuyekethisa okuphindelelayo kwezinhlelo ze-AI noma ukusetshenziswa okulinganisekayo kwe-AI ekuhlaselweni, lokho kungaqinisa icala elingokoqobo lokushintsha kwe-CSA kokubalulekile. Ngokuphambene, izinga lenhlolovo lilodwa alikwazi ukusungula ubungozi besikhathi esizayo, kufakazela ukuthi i-AI iyimbangela ehamba phambili yezigameko zamafu noma ibonise ukuthi izilawuli ezinconyiwe ziyasebenza ekusebenzeni.