Ukuhlakanipha okungekhona okwangempela ingashintsha kakhulu imakethe yemisebenzi eminyakeni ezayo, ikakhulukazi lapho i-cybersecurity iyathinteka. Noma i-AI ikhulula abasebenzi bomsebenzi oshubile, badinga ukuchitha isikhathi behlela amathuluzi abo. Abasebenzi badinga ukuqeqesha i-AI ukubona ukuhlaselwa okuba nzima futhi kube nzima ukugwema.
Iqiniso ngukuthi ayikho impendulo elula yokuthi ngabe i-AI izothatha isikhundla sabantu abasebenzi. Njengoba kuvela amathuba amaningi ukufezekiswa kwe-AI emabhizinisini, Kuzoba yimpahla yemikhakha ethile kanye nosongo kwabanye. Kunemiphumela eminingi ebukekayo ye-AI emphakathini, kodwa futhi nezinqwaba ezihle futhi.
Ososayensi bedatha kanye nabathuthukisi bahlala beza nezinhlelo zokusebenza ezintsha ze-AI. Kepha inhloso ihlala ifana: Ukuthuthukisa izimpilo zethu. Sesivele sibhekene nezinzuzo ezithile ze-AI ngokusetshenziswa kwayo okusakazekile kubasizi bezwi nezimoto ezizimele, Kepha kunokuningi okuzayo. I-AI izothinta umnotho kakhulu ngenxa yezicelo zayo ekuvuthweni kwe-cybersecurity. Ukuhlasela kwe-malware ne-rangaware kukhule ku-ubhubhane kulingana, kubangela ukulahleka okukhulu kwezezimali. I-AI ingalwa nalokhu kuhlaselwa ngokuqeqeshwa okufanele. Ngakho-ke kuzoba yimpi ende futhi enzima phakathi kokuhle nokubi.
I-AI ingasiza ukulwa ne-cybercrime ngokuthola amaphethini womsebenzi asayina okuthile akulungile. Kuyafana nokuthi ubuhlakani bokufakelwa busetshenziswa kanjani ezinsizakalweni zezezimali ukuthola inkohliso. I-AI ifeza lokhu ezinhlelweni ezizobhekana nezigidi zemicimbi njalo ngomzuzwana. Ngezikhathi ezinjalo zesiyaluyalu kanye ne-mayhem, Abakhwabanisi bavame ukuzama ukushaya. Kujwayelekile kuma-DDOs (Kusatshalaliswa Ukuphikwa Kwenkonzo) ukuhlasela.
Amakhono ka-AI wokuqagela enza kube lusizo olukhulu kulesi simo, yingakho, Njengoba sisondela 2022, Amabhizinisi amaningi azotshala imali kulobu buchwepheshe obunqenqemeni. Ngeshwa, Ama-Draftersters ayazi ngezinzuzo ze-AI. Ukuhlaselwa okusha kuyavela ukuthi sebenzisa ubuchwepheshe njengokufunda komshini. Lokhu kuhlasela kungadlula ekuvikelweni kwe-cybersecurity.
Ukufundwa kwemishini kusebenza kanjani ekulweni ne-malware?
Ukufundwa Komshini kuvame ukuqashwa embonini ye-antivirus ukukhulisa amakhono okuthola. Amasu wokufunda umshini Sebenzisa idatha yesampula ukudala imodeli yezibalo ebikezela ukuthi ngabe ifayili liphephile noma lingcolisiwe.
Ukufundwa komshini kuncike emaphethini amasha kuphela kunamazinga amakhodi ukuthola amaphethini ayingozi.
I-algorithm isetshenziswa kumaphoyinti wedatha abonakalayo amasethi amabili wedatha akhiwe ngesandla. Eyodwa iqukethe amafayela anonya kuphela namanye amafayela angewona amabi kuphela. Ngaphandle kokutshelwa ukuthi yimiphi izinhlobo zamaphethini noma amaphoyinti wedatha okufanele ubheke, I-algorithm yakha imithetho evumela ukuthi ihlukanise phakathi kwamafayela aphephile futhi ayingozi. Iphuzu ledatha yinoma yiluphi ucezu lolwazi mayelana nefayela, njengesakhiwo sangaphakathi sefayela, I-Compiler esetshenzisiwe, Izinsizakusebenza zemibhalo zihlanganiswe efayeleni, njalo njalo.
Isoftware ye-antivirus esebenzisa umshini wokufunda ingathola izinsongo ezintsha ngaphandle kokuncika kumasiginesha. Esikhathini esedlule, Isoftware ye-antivirus ngokuyinhloko ithembela ekuvumeni kweminwe, okuqhathanisa amafayela ku-database enkulu ye-malware eyaziwayo.
Abahloli besiginesha bangathola kuphela i-malware eyaziwayo engadala izingqinamba ezibalulekile. Kwenze isoftware ye-anti-malware isengozini enkulu kusukela emakhulu ezinkulungwane zezinhlobo ezintsha ze-malware athuthukiswa njalo.
Ngokolunye uhlangothi, Ukufundwa komshini wokuqeqesha ukuqaphela izinkomba zamafayili anonya futhi angewona amabi kungenzeka. Ivumela ukuba ubone amaphethini ayingozi futhi ubone i-malware - noma ngabe ibonwa ngaphambili.
Ososayensi banayo ukutholwa kwe-malware okuzenzakalelayo ngokudla idatha yangaphambi ye-malware esetshenzisiwe kwikhompyutha. Le ndlela ihlinzeka ngomshini ngokubuka okungaqondakali kwe-malware. Inethiwekhi ye-neural, Isihlahla Sokuthatha Isinqumo, nomshini wokusekela we-vector uyisebenzisele ukunquma impendulo efanele. Le ndlela inenzuzo ngoba ayithembeli ngezinto ezenzelwe ngesandla ezisuselwa kulwazi lochwepheshe lwesizinda.
Inethiwekhi ye-neural nayo ingaqeqeshwa kusetshenziswa izici kusuka ekuphukeni kwemibiko enonya okuhlelwa kabusha. Inethiwekhi yokuphakelayo, Ngokolunye uhlangothi, kuthatha njengokufaka uhlu lwemisebenzi engenisiwe namafayela e-DLL ahambisana nawo, kanye neMetadata kusuka kunhlokweni ephathekayo ephathekayo.
Ukuphakelwa phambili kanye nemiklamo yenethiwekhi ye-Neaural Aural, kanye nezimpawu zabo ezihlobene, zihlanganiswe kwinethiwekhi eyodwa ku-aural network esekelwe kwinethiwekhi. Ngemuva kokuhlanganisa imininingwane efundwe yimisebenzi yomibili ye-subnet, Le Nkulondla ihambisa okuphuma kokugcina. Inethiwekhi izokwazi ukuhlukanisa amacala esikhathi esizayo esebenzisa le ndlela ehlanganisiwe.
Ukufunda ngomshini kukude ne-Perfect
Ngenkathi ukufunda ngomshini kungaba yithuluzi elizuzisa kakhulu, yize kungenjalo ngaphandle kwezingqinamba. Omunye wamaphutha akhethekile womshini angcono kakhulu ukuthi awuqondi kabusha amamodeli azakhayo - nje awenze. Imane icubungule idatha futhi yenza izinqumo zisebenzise kahle kakhulu, Indlela efakazelwe ngokwezibalo.
Njengoba kushiwo ngaphambili, I-algorithm ihlinzekwa ngezigidi zamaphoyinti wedatha kepha akekho otshengisa ukuthi yimaphi amaphuzu wedatha ayizimpawu ze-malware. Yinto leyo imodeli yokufunda umshini okuzodingeka ukuthi azithole ngokwayo. Ngenxa yalokho, Akekho umuntu ongakwazi ngokweqiniso ukuthi imaphi amaphuzu wedatha angaphakamisa usongo, Ngokwemodeli yokufunda yomshini. Iphuzu ledatha elilodwa noma inhlanganisela eqondile ye 20 Amaphoyinti wedatha angasetshenziswa.
Umhlaseli ozimisele angathola ukuthi imodeli isebenzisa kanjani la mapharamitha ukuthola usongo futhi axhaphaza ubungozi. Yingakho abalingisi abahle bakwazi ukuhlukumeza ai. Njengoba beqaphela ukuthi isebenza kanjani, bangathola izikhala zokugwema ukutholwa. Umlingisi ononya angafaka ikhodi ehlanzekile etholakala kumafayili amhlophe amhlophe futhi aphazamise kumafayela azobe angena ngenye indlela futhi aphuze i-AI.
Izinhlelo zokufunda ngomshini zingafunda kakhulu njengoba idatha inikezwe yona. Imodeli ephumelelayo idinga okokufaka kwedatha enkulu, ngamunye wawo kumele abhalwe ngokunembe. Lawa malebula asiza imodeli ekuqondeni izici ezithile zedatha (e.g. ngabe ifayela lihlanzekile, -nolubi, noma okungafuneki).
Amandla wemodeli yokuqeqesha ngempumelelo ancike ekulebula okunembile kwedatha enikezwe yona. Kungaba nzima futhi kudle isikhathi ukufeza, nokho. Iphuzu ledatha elilodwa elihlanganisiwe lemininingwane phakathi kwezigidi zamaphuzu wedatha ahlukaniswe ngokuqondile kungaholela ekubhujisweni noma ngabe kubukeka njengephutha le-minuscule. Uma imodeli isebenzisa okokufaka okungafanele ukwenza isinqumo, Kungadala amaphutha. Lawa maphutha angaqhubeka aqhubeke namaphrojekthi wokuqeqeshwa esizayo. Kubangela umphumela we-snowball ongaba nemiphumela efinyelela kude ngokuzayo.
Ukuhlanganiswa kokudala nokusha kuyadingeka.
I-AI ayikwazi ukusinda ngokwayo. Indlela eyi-hybrid ye-AI ehlanganiswe ngezindlela ezazama futhi ezihloliwe njenge-software yokuqapha kude ingaba usizo. Lezi zinhlelo ziye zazungeza eduze 15 iminyaka ekanye. Ngakho-ke kunenqwaba yemininingwane ukukhomba impumelelo yabo.
Izinhlelo zokusebenza zokuqapha ezikude zinento yomuntu kubo ukuze zikwazi ukusiza emuva i-AI lapho kukhunjulwa ngenxa yokuhlukahluka okuthile. Ngokwesibonelo, Isoftware yokuqapha umakhalekhukhwini ingasetshenziswa ukulandelela abasebenzi ngomakhalekhukhwini kanye nokuqapha ama-imeyili e-Fishy ngesici sawo sokuqapha se-imeyili. Izindaba ezingamanga noma izihloko ze-Clickbait eziholela kumawebhusayithi anonya kungavinjelwa ngamakhono okulandela imithombo yezokuxhumana kwalezi zinhlelo zokusebenza.
I-XNSPY, Uhlelo lokusebenza lokuqapha lweselula olukude oluvezwe kuhlu oluningi lwezinhlelo zokusebenza zokulandela umsebenzi ezitholakala emakethe zinikeza amandla okuqapha ama-imeyili. Abaqashi bangasebenzisa lezi zinhlelo zokusebenza futhi bawafake kubasebenzi’ I-Android ne-iOS ukuvikela ukuhamba kwabo kokusebenza kanye nedatha yengqondo. Ngokubhekabheka kuleli khasi, Ungathola ukuqonda ngohlu lwezinhlelo zokusebenza zokuqapha zabasebenzi ezaziwa kakhulu.
Njengoba abezindaba zezenhlalo bequkethe izigidigidi zabasebenzisi nezindaba ezingezona amanga zishicilelwa njalo ngomzuzwana, Kuzothatha iminyaka ukuqeqesha i-AI ngempumelelo. Kepha njengoba abaqashi bangaxwayisa abasebenzi ukuthi bavule izixhumanisi zokukhwabanisa, Inethiwekhi yonke ingavikeleka nge-Remote Monitoring Izinhlelo.
Le software isetshenziselwa Ukulandelela Abasebenzi abanamaselula futhi ingabheka amathuluzi wokuthumela imiyalezo wokugcina wokuphela njenge-telegraph noma i-whatsapp, Futhi njengoba sibona umkhuba wezindaba ezingezona ezisakazwa ezingxoxweni zeqembu njenge-Wildfire, Lezi zinhlelo zokusebenza zingavimbela ukulahleka kwezomnotho nokuzezimali. I-AI ingaba nephuze kakhulu ukuvikela lezi ziteshi ngenxa yemvelo yazo eguquguqukayo futhi engalindeleki.. Ukubethelwa kokuphela kokuphela kusetshenziswa kuzinhlelo zokusebenza zemiyalezo eminingana, ukuhlinzeka abadlali abanonya umuzwa wamanga wokuphepha. Ithuluzi lokuhlola lemiyalezo esheshayo ye-XNSPY lingaphazamisa ukubethelwa kokuphela kokuphela kwezimo ezinjalo. Yona, ngakhoke, Ivikela iminyaka yocwaningo nokusebenza kanzima ekuweni ezandleni ezingalungile.
