{"id":14439,"date":"2026-05-20T17:25:03","date_gmt":"2026-05-20T15:25:03","guid":{"rendered":"https:\/\/sbunker.org\/?p=14439"},"modified":"2026-05-20T17:25:03","modified_gmt":"2026-05-20T15:25:03","slug":"inteligjenca-artificiale-arme-e-re-e-dezinformimit-ne-kosove","status":"publish","type":"post","link":"https:\/\/sbunker.org\/sr\/disinfo\/inteligjenca-artificiale-arme-e-re-e-dezinformimit-ne-kosove\/","title":{"rendered":"Inteligjenca Artificiale, “arm\u00eb” e re e dezinformimit n\u00eb Kosov\u00eb"},"content":{"rendered":"

P\u00ebrgatiti: Arjeta Krasniqi\u00a0<\/strong><\/em><\/p>\n

 <\/p>\n

N\u00eb vitet e fundit, Inteligjenca Artificiale ka marr\u00eb hov. Nga filtrat n\u00eb Facebook e TikTok e deri te chatbot-\u00ebt, IA po p\u00ebrdoret gjithnj\u00eb e m\u00eb shum\u00eb. Por nj\u00eb nga temat q\u00eb ka ngritur shqet\u00ebsim \u00ebsht\u00eb p\u00ebrdorimi i Inteligjenc\u00ebs Artificiale p\u00ebr krijimin e fotove dhe videove t\u00eb rreme, t\u00eb njohura si \u201cdeepfake\u201d.<\/p>\n

Problematike \u00ebsht\u00eb kjo teknologji po p\u00ebrdoret jo vet\u00ebm p\u00ebr arg\u00ebtim, por edhe p\u00ebr manipulim politik, dezinformim dhe krijim t\u00eb narrativave t\u00eb rreme, \u00e7ka e b\u00ebn\u00eb t\u00eb v\u00ebshtir\u00eb dallimin mes lajmit t\u00eb v\u00ebrtet\u00eb dhe atij t\u00eb fabrikuar.<\/p>\n

Me vet\u00ebm disa aplikacione dhe pak minuta koh\u00eb, mund t\u00eb krijohen video ku nj\u00eb person duket sikur po flet apo po b\u00ebn di\u00e7ka, e q\u00eb n\u00eb t\u00eb v\u00ebrtet\u00eb nuk ka ndodhur kurr\u00eb. K\u00ebto p\u00ebrmbajtje po shp\u00ebrndahen me shpejt\u00ebsi n\u00eb rrjetet sociale dhe shpesh po mashtrojn\u00eb mij\u00ebra njer\u00ebz, kryesisht n\u00eb periudha t\u00eb ndjeshme sikurse jan\u00eb zgjedhjet.<\/p>\n

Sipas analiz\u00ebs s\u00eb disa postimeve, p\u00ebrdorimi i inteligjenc\u00ebs artificiale n\u00eb krijimin e fotove dhe videove po ndikon gjithnj\u00eb e m\u00eb shum\u00eb n\u00eb m\u00ebnyr\u00ebn se si njer\u00ebzit e perceptojn\u00eb informacionin online. Nga rastet e analizuara del se p\u00ebrmbajtjet e manipuluara shp\u00ebrndahen shum\u00eb shpejt\u00eb n\u00eb dhe arrijn\u00eb mij\u00ebra p\u00ebrdorues brenda nj\u00eb kohe t\u00eb shkurt\u00ebr, ku shum\u00eb prej tyre u besojn\u00eb k\u00ebtyre publikimeve, pa e verifikuar burimin apo autenticitetin e tyre.<\/p>\n

P\u00ebrmes videove apo fotografive t\u00eb krijuara me IA, synimi \u00ebsht\u00eb t\u00eb deformohen deklaratat, t\u00eb d\u00ebmtohet reputacioni i personave publik\u00eb si dhe t\u00eb ndikohet opinioni i qytetar\u00ebve.<\/p>\n

Manipulimi q\u00eb synohet<\/strong><\/p>\n

Nj\u00eb nga faqet me nd\u00ebrveprimin m\u00eb t\u00eb madh, \u201cT\u00eb vdes\u00ebsh s\u00eb qeshurit\u201d<\/a>, publikon video dhe fotografi t\u00eb krijuara me inteligjenc\u00eb artificiale, t\u00eb cilat marrin mij\u00ebra shikime, p\u00eblqime dhe shp\u00ebrndarje nga p\u00ebrdoruesit n\u00eb Facebook.<\/p>\n

N\u00eb nj\u00eb video<\/a> shfaqet nj\u00eb djal\u00eb, i cili duket sikur \u00ebsht\u00eb duke e gdhendur n\u00eb dru figur\u00ebn e kryetarit t\u00eb LVV-s\u00eb, Albin Kurti, duke krijuar p\u00ebrshtypjen se b\u00ebhet fjal\u00eb p\u00ebr nj\u00eb punim artistik real. P\u00ebr ta b\u00ebr\u00eb t\u00eb duket m\u00eb bind\u00ebse, atij i \u00ebsht\u00eb veshur edhe fjalia: \u201cAlbin Kurti \u00ebsht\u00eb nj\u00eb udh\u00ebheqes i fort\u00eb dhe vizionar p\u00ebr Kosov\u00ebn prandaj po e skalit k\u00ebt\u00eb bas p\u00ebr t\u00eb nderuar pun\u00ebn e tij\u201d.<\/em><\/p>\n

Pamjet jan\u00eb gjeneruar n\u00eb m\u00ebnyr\u00eb q\u00eb t\u00eb duken autentike, sikur video \u00ebsht\u00eb incizuar n\u00eb nj\u00eb ambient pune apo studio artistike, ku personi \u00ebsht\u00eb i p\u00ebrqendruar n\u00eb procesin e krijimit t\u00eb skulptur\u00ebs. Megjithat\u00eb, n\u00eb realitet kjo ngjarje nuk ka ndodhur. Videoja \u00ebsht\u00eb e gjeneruar p\u00ebrmes inteligjenc\u00ebs artificiale, duke p\u00ebrdorur teknologji q\u00eb krijon skena t\u00eb rreme me pamje reale. Q\u00ebllimi i saj \u00ebsht\u00eb t\u00eb paraqitet si p\u00ebrmbajtje e v\u00ebrtet\u00eb p\u00ebr publikun duke e lart\u00ebsuar nj\u00eb politikan\u00eb t\u00eb caktuar.<\/p>\n

Kjo video ka marr\u00eb nd\u00ebrveprim t\u00eb madh n\u00eb rrjetet sociale, duke grumbulluar mbi 6.7 mij\u00eb p\u00eblqime, 358 komente dhe 226 shp\u00ebrndarje. Pjesa m\u00eb e madhe e komenteve jan\u00eb mb\u00ebshtet\u00ebse dhe pozitive, ku p\u00ebrdoruesit shprehin p\u00ebrkrahje, lavd\u00ebrime dhe besim ndaj figur\u00ebs q\u00eb skalitet n\u00eb dru. Reagimet e tilla tregojn\u00eb p\u00ebr ndikimin q\u00eb kan\u00eb p\u00ebrmbajtjet vizuale dhe videot n\u00eb form\u00ebsimin e opinionit publik dhe n\u00eb rritjen e angazhimit t\u00eb audienc\u00ebs n\u00eb platformat sociale.<\/p>\n

Tjet\u00ebr shembull \u00ebsht\u00eb nj\u00eb video<\/a> e publikuar nga faqja: \u201cKosova info +\u201c<\/a> me titullin: \u201cA po e fundos Vjosa Osmani politikisht Donika G\u00ebrvall\u00ebn?\u201d<\/em>. Aty shfaqet Vjosa Osmani dhe Donika G\u00ebrvalla n\u00eb videon e manipuluar me IA, me q\u00ebllimin q\u00eb t\u00eb paraqitet dram\u00eb politike dhe tension mes tyre.<\/p>\n

Rast tjet\u00ebr i manipulimit dhe dezinformimit \u00ebsht\u00eb nj\u00eb fotografi<\/a> me mbishkrimin: \u201cN\u00eb nj\u00eb emision n\u00eb Serbi e fal\u00ebnderon Vjosa Osmanin, Radoj\u00e7iqi \u2013 turp\u201d<\/em>. N\u00eb fotografi shfaqet Radoj\u00e7iqi, por p\u00ebrmes tekstit dhe kontekstit t\u00eb rrem\u00eb: \u201cshpejt\u00eb do t\u00eb m\u00eb shihni n\u00eb Prishtin\u00eb\u201d<\/em>, ku tentohet t\u00eb krijohet nj\u00eb situat\u00eb e paqen\u00eb dhe t\u00eb nxis\u00eb panik tek opinioni.<\/p>\n

M\u00ebsimdh\u00ebn\u00ebsi n\u00eb Fakultetin e Gazetaris\u00eb, Dren G\u00ebrguri, vler\u00ebson se rreziku m\u00eb i madh \u00ebsht\u00eb keqp\u00ebrdorimi i deepfake-ve p\u00ebrmes videove, fotografive apo audio-ve t\u00eb gjeneruara me IA. Sipas tij, qytetar\u00ebt shpeshher\u00eb nuk arrijn\u00eb t\u2019i dallojn\u00eb k\u00ebto p\u00ebrmbajtje dhe krijojn\u00eb bindje t\u00eb gabuara, sidomos n\u00eb koh\u00eb zgjedhjesh.<\/p>\n

\u201cKur shikojm\u00eb p\u00ebrmbajtje t\u00eb gjeneruara me inteligjenc\u00eb artificiale dhe analizojm\u00eb komentet n\u00eb rrjetet sociale, v\u00ebrehet se shum\u00eb qytetar\u00eb nuk arrijn\u00eb t\u2019i dallojn\u00eb ato nga p\u00ebrmbajtjet reale dhe i marrin si t\u00eb v\u00ebrteta. Kjo ka ndikim t\u00eb madh, sepse vendimet q\u00eb marrin qytetar\u00ebt shpesh bazohen n\u00eb at\u00eb q\u00eb shohin online, duke besuar se ajo \u00ebsht\u00eb realitet, edhe pse n\u00eb fakt mund t\u00eb jet\u00eb e fabrikuar. N\u00eb t\u00eb v\u00ebrtet\u00eb, b\u00ebhet fjal\u00eb p\u00ebr p\u00ebrmbajtje t\u00eb krijuar me AI, e cila mund t\u00eb ket\u00eb q\u00ebllime t\u00eb ndryshme \u2013 nga arg\u00ebtimi deri te keqinformimi apo manipulimi. N\u00eb disa raste, pasojat mund t\u00eb jen\u00eb serioze p\u00ebr qytetar\u00ebt\u201d<\/em>, thot\u00eb G\u00ebrguri.<\/p>\n

Problematike \u00ebsht\u00eb fakti se edhe mediume q\u00eb u dihet pron\u00ebsia<\/a> dhe jan\u00eb t\u00eb njohura n\u00eb sken\u00ebn mediale, p\u00ebrdorin fotografi t\u00eb krijuara me Inteligjenc\u00eb Artificiale p\u00ebr t\u2019i b\u00ebr\u00eb m\u00eb bind\u00ebse p\u00ebrmbajtjet q\u00eb publikojn\u00eb. Gazeta Demos<\/a>, n\u00eb nj\u00eb lajm t\u00eb publikuar ka p\u00ebrdorur fotografi t\u00eb gjeneruar me IA\u00a0 Glauk Konjufc\u00ebs dhe Enver Hoxhajt, p\u00ebr ta fuqizuar gjoja lajmin se ka ndodhur nj\u00eb takim mes tyre.<\/p>\n

Sipas G\u00ebrgurit, fotot ilustruese nga gazetar\u00ebt p\u00ebr t\u00eb krijuar imazh t\u00eb rrem\u00eb i cili nuk ka ndodhur n\u00eb realitet, nuk do duhej t\u00eb b\u00ebhen p\u00ebr shkak se ndryshon rrjedha e lajmit dhe narrativa q\u00eb u sh\u00ebrbehet qytetar\u00ebve.<\/p>\n

\u201cKjo praktik\u00eb \u00ebsht\u00eb problematike, sepse krijon nj\u00eb realitet q\u00eb nuk \u00ebsht\u00eb i v\u00ebrtet\u00eb. Mendoj se kjo nuk duhet t\u00eb vazhdohet nga gazetar\u00ebt tan\u00eb, pasi n\u00ebse fillojm\u00eb t\u00eb krijojm\u00eb situata me AI, at\u00ebher\u00eb n\u00eb nj\u00ebfar\u00eb m\u00ebnyre po cenojm\u00eb procesin e sakt\u00eb t\u00eb informimit gazetaresk. Brenda lajmit po futen elemente vizuale q\u00eb nuk kan\u00eb ndodhur dhe nuk kan\u00eb ekzistuar n\u00eb realitet, duke e deformuar k\u00ebshtu kuptimin e ngjarjes\u201d<\/em>, thot\u00eb G\u00ebrguri.<\/p>\n

\u00c7far\u00eb jan\u00eb deepfake?<\/strong><\/p>\n

Deepfake jan\u00eb foto, audio ose video t\u00eb krijuara apo t\u00eb manipuluara me inteligjenc\u00eb artificiale p\u00ebr ta b\u00ebr\u00eb nj\u00eb person t\u00eb duket real n\u00eb nj\u00eb situat\u00eb t\u00eb rreme. Teknologjia p\u00ebrdor fytyr\u00ebn, z\u00ebrin dhe l\u00ebvizjet e nj\u00eb personi p\u00ebr t\u00eb krijuar materiale q\u00eb ngjajn\u00eb me realitetin.<\/p>\n

N\u00eb fillim kjo teknologji p\u00ebrdorej kryesisht p\u00ebr arg\u00ebtim, meme apo efekte filmike. Por, sot ajo po p\u00ebrdoret edhe p\u00ebr manipulim, propagand\u00eb, mashtrime dhe sulme ndaj personave publik\u00eb apo qytetar\u00ebve t\u00eb zakonsh\u00ebm.<\/p>\n

N\u00eb koh\u00ebn kur inteligjenca artificiale po p\u00ebrdoret gjithnj\u00eb e m\u00eb shum\u00eb p\u00ebr krijimin e videove dhe fotografive t\u00eb rreme, \u00ebsht\u00eb e r\u00ebnd\u00ebsishme q\u00eb p\u00ebrdoruesit e rrjeteve sociale t\u00eb jen\u00eb m\u00eb t\u00eb kujdessh\u00ebm p\u00ebr at\u00eb q\u00eb shohin dhe shp\u00ebrndajn\u00eb. Shum\u00eb p\u00ebrmbajtje n\u00eb Facebook apo platforma tjera publikohen vet\u00ebm p\u00ebr t\u00eb marr\u00eb klikime, reagime dhe shp\u00ebrndarje, pa qen\u00eb t\u00eb v\u00ebrteta.<\/p>\n

Gjat\u00eb zgjedhjeve n\u00eb shum\u00eb vende t\u00eb bot\u00ebs jan\u00eb publikuar foto t\u00eb manipuluara t\u00eb politikan\u00ebve<\/a> q\u00eb gjoja kan\u00eb b\u00ebr\u00eb deklarata t\u00eb rreme<\/a>. N\u00eb disa raste jan\u00eb p\u00ebrdorur edhe z\u00ebra t\u00eb krijuar me IA p\u00ebr t\u00eb mashtruar.<\/p>\n

Disa shenja q\u00eb mund t\u00eb tregojn\u00eb se nj\u00eb video apo fotografi \u00ebsht\u00eb krijuar me inteligjenc\u00eb artificiale, jan\u00eb l\u00ebvizjet jo natyrale t\u00eb fytyr\u00ebs, z\u00ebri q\u00eb ting\u00ebllon artificial, cil\u00ebsia e \u00e7uditshme e figur\u00ebs apo mimikat jo reale. Edukimi mediatik dhe mendimi kritik jan\u00eb m\u00ebnyrat m\u00eb t\u00eb mira p\u00ebr t\u2019u mbrojtur nga manipulimet dhe dezinformatat. P\u00ebr k\u00ebt\u00eb arsye, verifikimi i informacionit, ndjekja e mediave t\u00eb besueshme dhe kujdesi para shp\u00ebrndarjes s\u00eb postimeve jan\u00eb tejet t\u00eb r\u00ebnd\u00ebsishme n\u00eb epok\u00ebn digjitale.<\/p>\n

Sipas Dren G\u00ebrgurit, p\u00ebrdorimi i inteligjenc\u00ebs artificiale n\u00eb pun\u00ebn e gazetar\u00ebve n\u00eb Kosov\u00eb ende \u00ebsht\u00eb n\u00eb nivel t\u00eb ul\u00ebt, rreth 42 %, q\u00eb \u00ebsht\u00eb duksh\u00ebm m\u00eb i vog\u00ebl krahasuar me vendet e tjera. Ai thot\u00eb se nd\u00ebr problemet kryesore \u00ebsht\u00eb mungesa e udh\u00ebzuesve dhe trajnimeve brenda redaksive, gj\u00eb q\u00eb e b\u00ebn p\u00ebrdorimin e AI-s\u00eb t\u00eb paorganizuar dhe kryesisht t\u00eb bazuar n\u00eb iniciativa individuale t\u00eb gazetar\u00ebve.<\/p>\n

Kjo situat\u00eb krijon rrezik q\u00eb gazetaret t\u00eb mos jen\u00eb t\u00eb p\u00ebrgatitur sa duhet p\u00ebr kufizimet dhe m\u00ebnyr\u00ebn e duhur t\u00eb p\u00ebrdorimit t\u00eb k\u00ebtyre veglave, duke e b\u00ebr\u00eb t\u00eb domosdoshme vendosjen e standardeve dhe edukimit profesional p\u00ebr p\u00ebrdorimin etik t\u00eb IA n\u00eb media.<\/p>\n

\u201cKjo n\u00eb vetvete mund t\u00eb krijoj\u00eb problem p\u00ebr shkak se gazetar\u00ebt mund t\u00eb mos jen\u00eb t\u00eb p\u00ebrgatitur mjaftuesh\u00ebm, t\u00eb trajnuar p\u00ebr k\u00ebto vegla t\u00eb inteligjenc\u00ebs artificiale se si p\u00ebrdoren ato dhe t\u00eb jen\u00eb t\u00eb vet\u00ebdijsh\u00ebm p\u00ebr limitet q\u00eb duhet t\u00eb ekzistojn\u00eb n\u00eb p\u00ebrdorimin e inteligjenc\u00ebs artificiale\u201d, thot\u00eb G\u00ebrguri.<\/em><\/p>\n

Sipas tij, IA nuk e rrezikon drejtp\u00ebrdrejt besimin n\u00eb media, por gjith\u00e7ka varet nga m\u00ebnyra se si redaksit\u00eb dhe gazetar\u00ebt e p\u00ebrdorin at\u00eb, duke e konsideruar IA thjesht si nj\u00eb veg\u00ebl mb\u00ebshtet\u00ebse por jo z\u00ebvend\u00ebsuese t\u00eb pun\u00ebs s\u00eb gazetarit.<\/p>","protected":false},"excerpt":{"rendered":"

P\u00ebrgatiti: Arjeta Krasniqi\u00a0   N\u00eb vitet e fundit, Inteligjenca Artificiale ka marr\u00eb hov. Nga filtrat n\u00eb Facebook e TikTok e deri te chatbot-\u00ebt, IA po p\u00ebrdoret gjithnj\u00eb e m\u00eb shum\u00eb. Por nj\u00eb nga temat q\u00eb ka ngritur shqet\u00ebsim \u00ebsht\u00eb p\u00ebrdorimi i Inteligjenc\u00ebs Artificiale p\u00ebr krijimin e fotove dhe videove t\u00eb rreme, t\u00eb njohura si \u201cdeepfake\u201d. […]<\/p>","protected":false},"author":18,"featured_media":14440,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[363,1242],"tags":[2582,2577,1829,624,2662],"ppma_author":[28],"class_list":["post-14439","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-disinfo","category-in-depth-story","tag-ai","tag-antidisinfo","tag-edukimi-medial","tag-kryesore","tag-manipulation"],"authors":[{"term_id":28,"user_id":18,"is_guest":0,"slug":"disinfo","display_name":"Disinfo","avatar_url":{"url":"https:\/\/sbunker.org\/wp-content\/uploads\/2023\/10\/Sbunker_FAKE_NEWS_001-min1647858578.jpg","url2x":"https:\/\/sbunker.org\/wp-content\/uploads\/2023\/10\/Sbunker_FAKE_NEWS_001-min1647858578.jpg"},"user_url":"","last_name":"","first_name":"Disinfo","description":""}],"_links":{"self":[{"href":"https:\/\/sbunker.org\/sr\/wp-json\/wp\/v2\/posts\/14439","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/sbunker.org\/sr\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/sbunker.org\/sr\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/sbunker.org\/sr\/wp-json\/wp\/v2\/users\/18"}],"replies":[{"embeddable":true,"href":"https:\/\/sbunker.org\/sr\/wp-json\/wp\/v2\/comments?post=14439"}],"version-history":[{"count":3,"href":"https:\/\/sbunker.org\/sr\/wp-json\/wp\/v2\/posts\/14439\/revisions"}],"predecessor-version":[{"id":14442,"href":"https:\/\/sbunker.org\/sr\/wp-json\/wp\/v2\/posts\/14439\/revisions\/14442"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/sbunker.org\/sr\/wp-json\/wp\/v2\/media\/14440"}],"wp:attachment":[{"href":"https:\/\/sbunker.org\/sr\/wp-json\/wp\/v2\/media?parent=14439"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/sbunker.org\/sr\/wp-json\/wp\/v2\/categories?post=14439"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/sbunker.org\/sr\/wp-json\/wp\/v2\/tags?post=14439"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/sbunker.org\/sr\/wp-json\/wp\/v2\/ppma_author?post=14439"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}